Black Holes Revisited

In August, the Kavli Institute for Theoretical Physics (KITP) at the University of California, Santa Barbara convened a conference called “Massive Black Holes: Birth, Growth and Impact.” The conference brought together a diverse group of scientists to discuss a range of new issues, including how the latest theoretical predictions fit with new observational data about supermassive black holes. On the meeting’s last day, The Kavli Foundation brought together four speakers to discuss the latest issues in the field, as well as look ahead to the biggest questions and challenges.

The participants:

• Andrea Merloni – Staff scientist at the Max Planck Institute for Extraterrestrial Physics in Garching, Germany.
• Priyamvada Natarajan – Professor in the Departments of Astronomy and Physics at Yale University.
• Tommaso Treu – Professor in the Department of Physics at the Univesrity of California, Santa Barbara.
• John Wise – Assistant Professor at the Center for Relativistic Astrophysics at the Georgia Institute of Technology in Atlanta, Georgia.

The following is an edited transcript of the discussion.

THE KAVLI FOUNDATION: This conference focused significantly on the role supermassive black holes may have played in the formation of the earliest galaxies and galaxy clusters. The conference also raised many interesting questions more broadly about how black holes grow. Now that the conference is over, what do you see as the biggest questions needing answers?

JOHN WISE: Because I'm focused on the formation of the seeds of black holes in the early universe, I think the biggest mystery is whether black holes are the endpoint of a massive star that has died or if they’re formed by the direct collapse of gas clouds that eventually grow into supermassive black holes. Also, I want to know what kind of galaxy forms around these seed black holes, and whether with the upcoming James Webb Space Telescope (JWST), we can actually detect them at extreme distances back to the early universe and discern the origin of their central black holes during the first billion years of the universe’s history.

ANDREA MERLONI: I am really interested in how black holes grow – how gas and other matter get in. It’s the matter falling into a black hole that gets heated to high temperatures and emits tremendous amounts of ultraviolet and X-ray light – and that’s the signature that tells us a black hole is there. We have a theoretical understanding of how matter gets into a black hole, forming a disk around it as it does so, for example. But we haven’t fully connected our theories to what we observe, especially with quasars, these incredibly bright centers of very distant galaxies that serve as beacons of the early universe.

TOMMASO TREU: I am very, very interested in the relationship between galaxies and black holes, kind of like what John was referring to. Like everybody I also want to see what the first galaxy was like, and understand all the relationships between black holes and early galaxies. It’s the classic chicken and egg problem – what was there first? Black holes or galaxies?

PRIYAMVADA NATARAJAN: At this conference it has become clear that we are missing something when trying to explain how a black hole grows into a supermassive black hole found in the center of a galaxy. We know that these supermassive black holes existed in the very early universe, some 12 billion years ago, because our telescopes have detected them. And for decades we’ve expected these supermassive black holes were all originally small black holes that somehow grew into these monsters at the hearts of galaxies. But our computer simulations of black hole growth, based largely on theoretical ideas, show that they prematurely stunt their own growth and fail to become the heaviest supermassive black holes that we are detecting.

So our theories need to be revised. One new idea is that the very first black holes did not form from dead stars at all; instead, they formed directly from the collapse of huge amounts of gas. I’ve been investigating this and it seems possible. If so, then during this formative period for our universe, this gas may have created black holes large enough to eventually evolve into our very first supermassive black holes.

But did it actually happen this way? That's the biggest question for me right now.

TKF: In pop culture, especially the movies, black holes are dangerous and destructive monsters that swallow everything up. But several talks at this conference have emphasized how they’re also great engines of creation and growth. Priya, how would you say our view of black holes has evolved?

NATARAJAN: Definitely for astrophysicists, especially in the last decade, black holes have moved from being an exotic object of marginal interest to being a necessary component in galaxy formation. Not long ago you would go to a conference on structure formation and evolution and no one would really care about black holes. They were curiosities at the centers of galaxies. But today it seems like black holes - given how much energy they emit - can actually be important for modulating galaxy assembly, by regulating star formation and generating hot gas reservoirs. This puts black holes front and center. They’ve become a key physical ingredient in all theories of galaxy formation.

MERLONI: At the same time we need to be careful. Today, black holes are being proposed as the magic that helps solve problems encountered in computer simulations of the development of large-scale structure – galaxies, galaxy clusters and dark matter. But we are realizing that black holes cannot really solve everything, and what we discover has an impact for the wider astrophysics community.

TKF: Is it possible to imagine the universe without black holes?

MERLONI: Cosmologists who perform computer simulations of the growth of large-scale structures such as galaxies cannot do them without black holes, so that is hard to imagine.

NATARAJAN: They really are inevitable. They're going to form, if our current theories work. They are the end products of stellar evolution, that’s for sure. Even if you don't buy any of the direct collapse seed formation ideas yet, black holes are inevitable. They are there, we have compelling evidence for the one in the center of our galaxy, and we also think they are going to merge and grow in mass. They will assemble into even more massive black holes, even if we start out with little piddly ones.

TKF: We know that many galaxy clusters are dominated by a massive central black hole that sucks gas toward it but at the same time emits huge amounts of energy as things fall in. Tommaso, what are your biggest questions about this process?

TREU: There are two main questions. First: how does the outflow of energy couple with the surrounding gas? We don't yet know the physics of that. Second, and the most interesting one for me, is how the outgoing energy affects the underlying distribution of dark matter. We want to know because you can learn about the physical properties of dark matter – whether it interacts with itself, whether it is warm or cold, and so on. But in order to understand that, you need to understand dark matter’s relationship to black holes – how black holes and dark matter “talk” to each other.

TKF: What are some of the current ideas about this?

TREU: One is that the black hole basically ejects mass in ways that are fast and what physicists call “nonreversible” – like big burps of gas. By doing this it changes the gravity around the black hole so rapidly, it changes the shape of the dark matter and depletes what otherwise would accumulate in the center of galaxy clusters. It's almost like you took half of the Sun and threw it out beyond Pluto, magically. All of a sudden all the planets orbiting the sun would find themselves in a very different gravitational environment.

TKF: Priya, you gave a talk at the conference on a new theoretical idea about how supermassive black holes may have been born. You mentioned this earlier, but tell us more about what’s behind this new idea.

NATARAJAN: We see bright beacons in the universe – quasars – in place powered by black holes that are roughly a billion times the mass of our sun in the young universe, just a billion years or so after the Big Bang. Growing these monsters so rapidly is a lot easier if the starting seed black holes were more massive than dead stars. Computer simulations of the formation of the first stars in the universe suggest that black holes arising as the endpoint of stars cannot explain how some grow into supermassive black holes at the centers of galaxies – which we know happens. Black holes have to start out much larger initially in order to grow into the supermassive black holes at the centers of galaxies that we see today, unless we are missing some fundamental physics in our understanding of gas accretion. And the direct collapse of gas would give them that bigger starting point – bigger than if they are produced only from the extinguished cores of stars.

"It’s the classic chicken and egg problem – what was there first? Black holes or galaxies?" —Tommaso Treu

TKF: So how do researchers think this “direct collapse” of gas worked to form the first black holes?

NATARAJAN: What we think happened is that in some sites in the universe where early galaxies assembled, massive gas disks formed generating inflows of gas toward the center. And this runaway accumulation of gas to incredibly high densities led directly to the formation of massive black holes – and not stars. Instabilities in the inflowing gas prevented the fragmentation of the gas, which we believe is needed to form stars. But at the same time these instabilities drove the gas inward to form these direct-collapse black holes. So in these rare sites in the universe the formation of stars was halted and black hole seeds formed instead. Exactly how it proceeds in detail, we don't currently understand. There may be this object generated called a quasi-star that has a big gas reservoir and a trapping envelope around it, so that you end up with something that is about 1,000 times to 10,000 times the mass of the sun.

TKF: How big is the debate about this?

NATARAJAN: It is a big one. We think from our earlier work there are some observational signatures to support the idea, but they're pretty hard to confirm. This is why I am so excited about JWST, because we'll be able to see back to the time when we think the earliest black holes assembled. For me, the biggest challenge is really understanding the physics of those massive seeds and trying to understand how to detect them. There has to be more than one way of collapsing gas and actually making a black hole. In the end, it may turn out that in the early universe there are sites that formed the first stars, and other regions where black holes were created from the direct collapse of gas. But I personally think that it's not going to be either/or. We may be able to distinguish these possibilities because there are unique fingerprints that each of these channels leaves behind. Direct collapse seeds will produce galaxies that have an overly-massive black hole seed compared to the stars for a period of time. This class of galaxies, which we refer to as OBGs (Obese Black Hole Galaxies) are detectable by JWST which is really exciting.

TKF: And once these first black holes did appear, how do we think they grew into supermassive black holes, which are found at the centers of galaxies and galaxy clusters?

NATARAJAN: You start with some initial mass, and then our current theoretical understanding and models say you grow because you accrete gas. And you also grow because you actually merge with another black hole. It’s the combination of both, with the accretion being the dominant player.

MERLONI: There may be some places in the universe where galaxies rarely interact with each other, and thus black hole mergers are not driving their growth, but those are quite rare.

"One new idea is that the very first black holes did not form from dead stars at all; instead, they formed directly from the collapse of huge amounts of gas. I’ve been investigating this and it seems possible." — Priyamvada Natarajan

NATARAJAN: Black holes grow through the accretion of gas, but we also know that galaxies merge and so the central black holes at the center of each galaxy must also merge. The picture that people have been pushing so far is that the process of the merger of galaxies will trigger a gas accretion episode. You have a “feast” when you have a merger, a black hole feast. They merge and become one, but the majority of the mass growth is not from the merger; it’s from the feast they have from gas coming in during the merger. So mergers trigger accretion, or feeding, episodes of the gas. And that's how we thought you could build up a supermassive black hole.

However, as we’ve built up a population of known black holes across cosmic time, it’s become clear that you probably need both these dramatic feasting episodes and some steady ordinary low-level trickling in of gas over cosmic time.

TKF: John, what do computer simulations say about this?

WISE: We've simulated the growth of black holes beginning with early galaxies that are much smaller than the Milky Way. What we’ve found is that in the early stages, these black holes have a really hard time growing because the energy they emit – mostly in X-rays – heats up and puffs up the gas surrounding the black hole. So you're basically burning off all your fuel that you would otherwise accrete. But I think that’s only in the initial stages. This is eventually offset by the many black hole mergers and “feasts” that Priya talked about that occur during the first billion years. Then once a supermassive black hole is embedded at the center of a fully realized galaxy that can actually sustain star formation, cold gas is funneled toward the center of the galaxy and feeds the supermassive black hole and its growth – rather than mergers and feasts.

TKF: Andrea, how large can a supermassive black hole grow?

MERLONI: The upper limit, we think, is about 10¹⁰ solar masses. We’ve seen a supermassive black hole of that mass at the center of a galaxy cluster. We also know the size of a supermassive black hole at the center of a galaxy, or galaxy cluster, correlates quite closely with the overall mass of that galaxy or cluster. But we still don’t know why.

WISE: I always find it interesting to find galaxies without a central supermassive black holes. One example of this is M33, the Triangulum Galaxy and a member of our local group of galaxies. No one’s ever detected a central black hole there.

MERLONI: It may be too small to have been detected, we don’t know.

TKF: Given the questions raised at this conference, what future observational missions do you think might answer some of them?

NATARAJAN: JWST would have the biggest reach for observing extremely distant black holes. And because of the particular filters on it, JWST will have the ability to uncover these black holes obscured by dust, which is what we expect to see in the early universe. For some of the other questions we still need other instruments, such as how black holes grow by accreting gas. A wide-field X-ray mission would be valuable, but the United States doesn’t have anything planned right now.

"As scientists, what we do is try to understand phenomena and put the pieces together. It's almost guaranteed that if you spend your time on black holes, you are bound to find new things and surprises." — Andrea Merloni

MERLONI: We are actually working on that. At the Max Planck Institute for Extraterrestrial Physics in Germany, we are building an X-ray telescope to survey the entire sky looking for active black holes. No objects other than black holes emit such copious amounts of X-ray radiation, and we expect to find about three million of them. The telescope is called eROSITA, which stands for extended ROentgen Survey with an Imaging Telescope Array. It will fly as part of a joint Russian/German mission called SRG, scheduled for launch at the end of 2014.

To find the most distant black holes, the next big mission planned is called Athena+. It will be an X-ray space telescope, and it’s expected to be launched in 2028 by the European Space Agency. Its goal is to discover black holes more than 13 billion light years away, corresponding to a time when the universe was 460 million years old. Athena will also have a wide field view, but it will go much deeper than eROSITA.

TKF: And John, what will be a game-changing advance for your computer simulations?

WISE: Over the past five years, GPUs, or Graphics Processing Units, have really pushed the field. These are basically graphics cards that are used commercially in video games. But scientists can use them for general purpose computing. The top five supercomputers in the world all have these GPUs now.

TKF: So with these GPUs, what is your dream simulation?

WISE: I would like to tie together the small-scale physics of a black hole accretion disk all the way up to the physics that governs an entire galaxy. That's far-reaching right there, and maybe 30 years down the line we can connect it to the cosmological scales of galaxy formation. To bridge all the gaps in our knowledge today – that’s what I’d like to do. Maybe by 2043 we might even have quantum computing to help us. But I’ll almost be retired by then.

Today, though, it is becoming easier and easier to use these processors. In years past you had to translate your code and use a different programming language – and astrophysicists are not computer scientists. Now, they're making it easier to write code that is compatible for these GPUs. In some sense, we are dependent on the software companies.

"I would like to tie together the small-scale physics of a black hole accretion disk all the way up to the physics that governs an entire galaxy." — John Wise

TKF: Is the reality of what we’re learning about black holes stranger than any movie or fictional account?

NATARAJAN: For me, black holes remain as enigmatic in fact as they do in fiction.

MERLONI: As scientists, what we do is try to understand phenomena and put the pieces together. It's almost guaranteed that if you spend your time on black holes, you are bound to find new things and surprises.