Bookmark and Share

A Conversation with George Efstathiou

WHEN HE WAS ABOUT TEN YEARS OLD, cosmologist George Efstathiou faced a seemingly easy question: “Why is the sky dark at night?” He soon learned that the question wasn’t so easy to answer, and that profound insights often reveal themselves through the simplest of questions.

George Efstathiou, director of the Kavli Institute for Cosmology at the University of Cambridge (Courtesy: G. Efstathiou)George Efstathiou, Director of the Kavli Institute for Cosmology at the University of Cambridge (Courtesy: G. Efstathiou)

Dr. Efstathiou, director of the Kavli Institute for Cosmology at the University of Cambridge, has pursued a life in science asking some of the biggest and most fundamental questions in nature: How did the universe begin? Where do stars and galaxies come from? What can we learn from the oldest light in the cosmos? The answers, and the mysteries that remain, have transformed the way we see the universe and our place in it.

In October, Dr. Efstathiou received the 2013 Nemitsas Prize in Cyprus, a prestigious award honoring Cypriot citizens and those with Cypriot heritage for their accomplishments in science, medicine, engineering and the arts. Dr. Efstathiou was honored in part for his contributions to our understanding of one the biggest puzzles in the universe: dark matter, the mysterious stuff that makes up 85 percent of all matter in the universe but still hasn’t been directly seen.

Dr. Efstathiou also was recognized for his leading role in the Planck satellite mission, which has studied the cosmic microwave background, the relic radiation left over from the Big Bang nearly 14 billion years ago and the oldest light in the universe. This background radiation fills every part of the sky, and temperature variations from one part to another represent variations in the density of matter that existed in the very early universe. And it’s these differences in density from which all structure in the universe grew – including gas, stars, galaxies, clusters of galaxies and the dark matter that holds it all together. By studying these details, Planck has opened an exquisite window on what the universe was like shortly after the Big Bang. 

In a recent interview with The Kavli Foundation, Dr. Efstathiou reflected on his latest work, the first sparks that fired his imagination about cosmology, and other subjects.

The following is an edited transcript.

THE KAVLI FOUNDATION: In October 2013 the Planck spacecraft was turned off, marking the end of its 4.5-year mission to measure tiny temperature fluctuations in the cosmic microwave background in unprecedented detail. As a result, we now have the sharpest image ever of the universe when it was only 380,000 years old. How do you think cosmologists will regard Planck’s accomplishments in coming decades?

GEORGE EFSTATHIOU: Well, I wish I knew. It’s important to remember the results that we released back in March came from a very conservative analysis of the temperature fluctuations. So we didn't really push the data very hard, although what we reported was more precise than any previous experiment. These latest results fit with our basic model for cosmology, which includes two defining features. One is a rapid, inflationary expansion of the universe a fraction of a second after the Big Bang. The other is that cold dark matter – exactly what it is remains a mystery – gravitationally dominates all other matter in the universe and binds together galaxies and clusters of galaxies.

I think there's no denying that we've uncovered a fundamental truth about the universe. Our basic cosmology model is correct, in that what we actually see supports our theories for both inflation and also for the existence and extent of dark matter. We now know that any deviations from this basic cosmological model are small and will be hard to measure.

TKF: What comes next for the Planck’s European-led science team, which includes you and nine other lead researchers at institutions across Europe and in the United States?

EFSTATHIOU: We’re about to begin the next phase of analysis, and this will be a much more complicated analysis of the data. During this process, we have to decide whether there is any evidence for new physics – physics that is not included in our current theories for the origin and evolution of the universe. The stakes are very high, because anything that we find must be repeatable, pass internal consistency tests and so on before we can make a convincing case. Until this next phase of analysis is finished in June 2014, we don't really know what Planck’s legacy will be.

TKF: Where do you think the Planck mission stands among the many projects you've worked on? And how will it continue to influence your work?

EFSTATHIOU: Planck has been the biggest and time-wise longest of the large projects I’ve worked on. I’ve been involved in the mission since 1992 – more than 21 years. It definitely has been the most challenging. There will be three more years of work on Planck, at least.

Planck DataThis is an image of the oldest light in our universe, imprinted on the sky when the universe was just 380,000 years old. It was obtained by measurements made with the European Space Agency's Planck spacecraft. The image of this radiation, called the Cosmic Microwave Background, shows tiny temperature fluctuations that correspond to regions of slightly different densities, representing the seeds of all future structure: the stars and galaxies of today. George Efstathiou has been a leading scientist with the Planck mission since 1992. (Credit: ESA, Planck Collaboration)

Now that we have such precise data from Planck, we have a comprehensive view of the very early universe. But we don’t have data that is equally precise from many other areas of cosmology. What that means is that the quality of data from the very early universe, compared with other time periods, is mismatched. Many other observations – of distant supernovae, for example – are just not as clean and the quantity of data for them is not as high. And the less data we have, the more susceptible we are to systematic errors. But from the cosmic microwave background, we have these beautifully precise measurements. Of course, there are important physics questions – such as the nature of dark energy – which you can't answer just from observations of the cosmic microwave background.

So in the future, I can use my experience with Planck and other projects I’ve worked on to look more critically at other types of data – to see whether we can make sense of it given what we know from Planck.

TKF: Let’s take a look back for a bit. You were born and raised in the U.K. but your parents immigrated to England from Cyprus in the early 1950s. Tell me about that, and what it was like to receive an honor from your parents’ home country.

EFSTATHIOU: My father was 18 when he came to England in 1950, and my mother came over from Cyprus a couple of years later, when she was 14. They immigrated at a time when people were needed in England to help re-build the country, and there were opportunities for work there that just didn’t exist then in Cyprus. I was born in London, the oldest of three siblings. I was brought up speaking Greek, and today I have knowledge of the language at sort of the household level.

The trip to Cyprus in October was fantastic. My parents came, and because it was a Cypriot prize it had another layer of meaning. It was something that my parents could relate to, as opposed to some of the other physics prizes I’ve won over the years. It was a lot of fun also for my two youngest kids.

"...[W]e still don’t understand what most of the stuff in the universe is at all. We have a theoretical model for how the universe began and how it has evolved, but we’re just scratching the surface. There is a lot to be done..." – George Efstathiou

TKF: What first sparked your interest in science?

What got me into science was actually my father. He bought an encyclopedia and they were the only books in the house, and I just started from an early age plowing through this stuff. What grabbed me were the articles dealing with the big questions. How do planets form? How do stars form? What do we know about the universe?

TKF: Did you have a telescope as a child?

EFSTATHIOU: I got a telescope when I was about 11, a 4 ½-inch refractor. I did time tests on the satellites of Jupiter, tracking their orbits around the planet. I looked at the phases of the moon, and made drawings of its surface.

TKF: Some adults who grew up in the 1970s and ‘80s point to Carl Sagan as someone who turned him or her on to astronomy. Who did this for you in the 1960s?

EFSTATHIOU: I remember being really strongly influenced by Patrick Moore, the amateur astronomer and host for “The Sky at Night.” The BBC show was very influential with a lot of people who grew up in the U.K. I remember one time – I was about nine or ten years old – I wasn't really paying attention to the TV and he came on and said, “Well, today I'm going to tell you why it's dark at night.” And I just stopped, and I said, “Oh, come on. That is really easy. It's obvious why the sky is dark at night.”

So then he described Olbers’ Paradox. He explained that in an infinitely large and infinitely old universe we should see a star along every line of sight, and stars are very bright and so the sky should be blazing. But of course it isn’t, and the reason – the explanation for the paradox – is that the universe is not infinitely old. Because light has a finite speed and the universe is not infinitely old, some starlight hasn’t yet reached Earth. Added to this is the fact that the universe is expanding, and so the energy of very distant starlight is reduced by the time it reaches us. It’s shifted into the infrared part of the light spectrum, and it becomes invisible to our eyes.

I remember I was just really shocked when I heard this for the first time. Suddenly my youthful arrogance was demolished.

TKF: There are rarely simple answers to big questions.

EFSTATHIOU: The fact that you can draw such far-reaching conclusions from straightforward observations – it’s one of the great aspects of science. There is profundity behind the simplest questions. Being able to ask questions that require really, really deep insights into the way the universe works to get solutions is just fascinating. It's still fascinating.

The Efstathiou FamilyThe Efstathiou family. A young George stands between his parents, Petros and Christina Efstathiou. (Courtesy: G. Efstathiou)

TKF: So Patrick Moore had a big influence on you, but you also grew up during the heady days of the American space program. What impact did that have on you as a child?

EFSTATHIOU: Massive, massive. It was just inspirational. I stayed up to see Neil Armstrong, but even more, to see Apollo 8 going around the moon just seemed unbelievable. In hindsight, one of the things that was so amazing about the Apollo program was just the rate at which they were doing things. The speed with which it was done. It was really astonishing. Everything was done on this accelerated schedule, so as a child it meant that there was always something new happening that was just awe-inspiring.

TKF: When did you know that cosmology would become your life’s work?

EFSTATHIOU: I went to Oxford as an undergraduate, and while I was there I actually became really interested in particle physics. During my final year I wasn't really sure whether to continue in particle physics or cosmology. We had a very famous particle physicist then named Richard Dalitz, a leading theoretical particle physicist at the time, and I went to see him. He said basically that it was a no-brainer: progress in particle physics at that time had slowed and cosmology was wide open, and I should do cosmology. And so I did. That really was the most influential advice I had at the time.

TKF: So there was more opportunity at that time for a young researcher to make an impact in cosmology?

EFSTATHIOU: One way to look at it is to consider the Higgs boson, which was discovered in July 2012 and made particle physics major news. But it was back in the 1960s that Peter Higgs, Francois Englert and Robert Brout independently proposed its existence, and that it imparts mass to other subatomic particles. It took half a century to find evidence for this. Cosmology, on the other hand, was wide open when I started, and progress in the field over the course of my career has been tremendous. At the same time, we still don’t understand what most of the stuff in the universe is at all. We have a theoretical model for how the universe began and how it has evolved, but we’re just scratching the surface. There is a lot to be done, and there is the potential for revolutionary new things. The history of the subject is that nature throws up stuff that we could not have dreamed of.

TKF: In 2011, you won the Gruber Prize for developing computer simulations back in the 1980s that helped lead to the idea that cold dark matter drives the way galaxies and galaxy clusters evolve. What focused you on this work back then, and what were you trying to solve?

EFSTATHIOU: Back in the 1980s, leading cosmologists like James Peebles had started to become interested in developing computer simulations that could model how galaxies cluster over cosmic time. This required developing programs that accounted for Newton’s laws of gravity, which describe how ordinary objects interact with one another.

"The fact that you can draw such far-reaching conclusions from straightforward observations – it’s one of the great aspects of science. There is profundity behind the simplest questions." – George Efstathiou

People were doing numerical simulations using only about 1,000 particles because computers were slow. However, I had heard about a group at the University of Reading in the U.K. that was developing computer codes to do problems related to solid-state physics. For example, finding the melting points of crystals and using various assumptions about what’s involved.

What I realized is that the techniques they were developing could be applied to cosmology. These techniques could be generalized, and used very efficiently to simulate the interactions among large numbers of particles. So I contacted them and spent some months of the University of Reading, and then came back and wrote a code that allowed me to instantly go to simulations with 20,000 and 30,000 particles. At the time, it was revolutionary.

TKF: What kinds of roadblocks did you run into at that time?

EFSTATHIOU: I was a graduate student, and to get the computer time was very difficult. And nobody was supervising this work so I had to persuade public bodies here in the U.K. to give a student a large allocation of computing time on what was then the fastest computer outside of restricted defense laboratories. So that was the major roadblock, and then there was the sheer tedium of trying to run these computer simulations. Most of the work that I did for my thesis was done over Christmas and Easter vacations, because that’s when the computers were available.

TKF: At what point did you know you were on the right path, that you really had a breakthrough?

Dark Matter DistrubutionThis NASA Hubble Space Telescope image shows the distribution of dark matter in the center of the giant galaxy cluster Abell 1689. Dark matter is an invisible form of matter that accounts for most of the universe's mass. Hubble cannot see dark matter directly. Astronomers inferred its location by analyzing the effect of gravitational lensing, where light from galaxies behind Abell 1689 is distorted by intervening matter within the cluster. In the 1980s, George Efstathiou developed computer simulations that helped lead to the idea that dark matter drives the way galaxies and galaxy clusters evolve. For this work, he won the prestigious Gruber Prize in Cosmology in 2011. (Credit: NASA/JPL-Caltech/ESA/Institute of Astrophysics of Andalusia, University of Basque Country/JHU)

EFSTATHIOU: When I went to Reading and saw the codes and started playing with them. I knew right away it would be a straightforward path to adapt it to cosmology – and to scale it up to large numbers of particles. And it worked brilliantly well, and so I began a collaboration with Carlos Frenk, Marc Davis and Simon White. This was during a time when inflationary theory was being developed for the first time, and when researchers were thinking that dark matter might be some kind of elementary particle. So that’s when we did the first simulations. But by the mid-1980s, I didn't want to stay in that field.

TKF: Why?

EFSTATHIOU: I didn't think it was intellectually challenging enough. You could refine numerical techniques and you could do new calculations with each increase in computing power, but I wanted to do more physics.

TKF: These computer simulations have obviously grown much more sophisticated since the 1980s. But what limitations do you think persist today?

EFSTATHIOU: This is also related to why I dropped the subject. Simulating the evolution of dark matter in the universe is purely a gravitational problem, and that problem is solved. We can do simulations now at a very high resolution. Simulations in cosmology have shifted toward studying the formation of galaxies and the behavior of gas – that is, how gas condenses into halos and forms into stars and galaxies. The difficulty with this is that it's not possible to solve this problem from first principles. The physics is just too complicated, so people have to put rules in for simulations – rules about how they think stars form, how the energy injected from stars affects the surrounding gas, and so on. These are prescriptions, and the results are very, very sensitive to these prescriptions. There's a lot of work and progress being made, but it's not a clean problem. And it's not a problem that can be easily solved with faster and faster computers. The dynamic ranges are too great.

TKF: In other words, how stars form and so on is not fully understood, and so the data that goes into these simulations aren’t of the highest quality.

EFSTATHIOU: That's right. I'm not trying to be critical of people who are working in that field. But my own personal point of view is I like to work on problems where there's a potential for making scientific progress more reliably and faster.

TKF: How much of your own work do you still do on a chalkboard? Is there still intellectual work for which you don’t need a big computer?

EFSTATHIOU: Quite a lot. Obviously not a lot during a big project like Planck. I‘ve structured my own science career around big projects, but sometimes there are interesting science problems that you can figure out theoretically. These are sort of “ideas” papers, if you like. In cosmology you can still do that. If you have an idea, it might take you only two or three days to figure it out and write it up.

TKF: Many people say we’re living today in a kind of “Golden Age of Cosmology.” Why is that?

EFSTATHIOU: Just over my own career, cosmology has been completely transformed. When I started as a graduate student, people were still arguing over evidence for a hot Big Bang. We had no idea where structure in the universe – galaxies and clusters of galaxies – came from, or what the composition of the universe is. That's all happened during my career. However, we still don’t have a clue what 95% of the universe is. So, the big question is, “How is it going to develop?” and that’s impossible to predict. Nature will throw up surprises. That's been the history of the subject.

— Winter, 2013 (Interview conducted by Bruce Lieberman)