Gravity's Engines: How Bubble-Blowing Black Holes Rule Galaxies, Stars, and Life in the Cosmos

by Scharf, Caleb

  If the skies are clear tonight, go outside and take a look around. Perhaps the Moon is visible. But what you see is not the Moon, but the past Moon. It’s the Moon as it was 1.3 seconds ago. Up there in the sky is another object, tiny but brilliant. It’s the planet Jupiter. Or rather, it is Jupiter as it was forty minutes ago. Look around a little more and find the bright stars. If you can see the southern sky, one of the brightest stars of all is Alpha Centauri A, as it was a little over four years ago. Other bright stars are glowing the way they were several decades ago. If the night sky is dark enough you can make out the hazy splash of the plane of the Milky Way galaxy, the projected light of the nearest spiral arms. Most of the light you see has been traveling toward you for thousands of years.

  Because of photons’ limited speed, we’re forever trapped by time, blanketed and shielded from whatever is happening in the cosmos right now. But the truth is that it is our minds that have a problem. We need to let go of our conceit that we really witness anything “as it happens.” When I drop a coin, I see it hit the ground a few nanoseconds after the coin “thinks” it does. If I watch a seagull scooping up a fish from the distant ocean, the gap in time between its hungry gulp and when I witness it can be tens of nanoseconds. It takes that long for the light reflected from these objects to reach me. The simple fact is that when events occur is all relative, something that is of course deeply embedded in Einstein’s descriptions of the physical universe. Luckily this characteristic of nature also provides us with the means to practice cosmic paleontology.

  In 1962, when astronomer Maarten Schmidt discovered how to interpret the light coming from a distant quasar, he realized that those photons had been en route for 2 billion years. Since that extraordinary measurement, astronomers have striven to push further and further back in cosmic time. We’ve looked for supernovae, quasars, radio-emitting galaxies, ordinary galaxies, and clusters of galaxies at ever-increasing distances. At times the quest has been highly competitive. Scientists highlight the cosmic distance of a new discovery front and center in the title announcing their work, clearly claiming bragging rights. The following week another team of researchers will try to one-up that discovery by perhaps 100 million years’ worth of intergalactic time. Astronomers take pride in their ability to eke out new objects that are fainter and more difficult to characterize than anything that came before. Indeed, it’s the nature of the subject, an indelible part of its history and practice.

  The result of all these efforts is a cosmic time line. Just like fossil hunters carefully brushing away the dusty grains of rock entombing a specimen, astronomers peel away layer upon layer of spacetime strata. In doing so, we can track the ways in which the populations of stars and galaxies have evolved as the universe has aged. It is how we know that quasars, the most massive and most actively feeding black holes, used to be far more prevalent. It is also how we know that the populations of stars and galaxies of recent times have changed from their earliest days. But why all this galactic evolution happens is still a major question. Theoretical astrophysicists use sophisticated computer simulations in their efforts to understand the steps involved. These virtual worlds incorporate the effects of gravity and gas pressure, and even attempt to model the nature of star formation and the complex interplay of energy and matter in different environments. Yet again and again many of these simulations have produced overgrown galaxies, systems the likes of which are nowhere to be seen in the universe. They contain too many stars, far more than really exist. This is a huge problem, and tells us that we are missing something, some piece of astrophysics that is not yet properly contained in our theories. But we’re making progress, and this progress is part of the story here. The bubbles blown by supermassive black holes are one critical component of the answer, but there is still much more at play that we need to understand.

  It’s in “today’s” universe that we’ve clearly seen the dramatic relationships between supermassive black hole energy output and the nature of galaxies and galaxy clusters. But we need the entire cosmic time line of this phenomenon in order to understand fully how the universe got to be the way it is. Following the fossil clues from the deep past is going to help give us the answers. So let’s now take a look at the story of one particular very distant, very strange, and very revealing place in the young universe.

  *

  My entry into the science of supermassive black holes came about because I was interested in something that I thought had nothing to do with these objects at all. For several years I had been pursuing clusters of galaxies across the universe, together with several colleagues around the world. Not literally, of course. Sadly, we were not superheroes. We chased them down the way astronomers chase down any cosmic objects: by surveying the skies with new tools and new persistence. More specifically, we were beachcombers, sifting through the sands of a large database of astronomical X-ray imagery. It all came from the orbiting X-ray telescope called the Roentgen Satellite, or ROSAT for short, a joint European and U.S. space mission. The mission was named after Wilhelm Roentgen, a German scientist distinguished by being the first person to receive the Nobel Prize for physics in its inaugural year of 1901. He was the discoverer of the mysterious “X-rays,” or Roentgen rays, that we are now familiar with at our dentist’s offices and hospitals. These were produced as a side effect during his laboratory experiments with cathode rays—beams of electrically accelerated electrons. He noticed that despite placing a thin aluminum screen, and then a sheet of cardboard, across the end of his experiment, something was still getting through and producing fluorescence in nearby materials. He called this unknown phenomenon “X-rays.” Roentgen was a talented scientist, but little did he know at the time the role that this radiation plays throughout our universe.

  During the early 1990s, ROSAT had taken long-exposure digitized images of X-rays emanating from all manner of astrophysical objects—thousands, in fact. As is common practice with almost any large telescope or astronomical instrument, astronomers petitioned the organizations running ROSAT about their favorite things to look at, some known and some exploratory. The telescope would eventually make the observations for the lucky few whose scientific arguments won out in a process of review by their peers. Later on, all this data was placed into a great electronic repository, archived for anyone to use. Scavengers like us could then trawl through the raw material to look for gemstones among the rough.

  We were a motley crew. Three of us, myself, the English astronomer Laurence Jones, and the American astronomer Eric Perlman, all ended up spending a couple of years at NASA’s Goddard Space Flight Center in suburban Maryland just outside Washington D.C. Jones and Perlman were the first to begin sniffing around in the piles of data from ROSAT, along with Gary Wegner, an astronomer at Dartmouth College. I joined in through my related work on the large structures in the universe, what would later be known as the cosmic web. And, although he was thousands of miles away in Hawaii, we also teamed up with Harald Ebeling, a German-born astronomer with a knack for clever computer algorithms, to sieve through X-ray telescope data looking for interesting objects. Equipped with computers and physics in place of hard hats and lanterns, we were cosmic data miners.

  Our goal was to try to find new and ever-more-distant examples of clusters of galaxies. Our clues came from the hot and tenuous gas harbored by the gravity wells of these huge structures. This was the same gas that would later be seen to contain the bubble-blowing black holes in our nearby universe. We looked for the wide, fuzzy smears of X-ray photons from this hot material that had been captured serendipitously in the ROSAT data archives. Once we found these features, we inspected their locations using earthbound telescopes to detect and count the galaxies that might be there. Clusters are as their name implies: their galaxies group together, making a crowded patch on the sky. Many X-ray smears turned out to be just stars or galaxies superimposed by chance, but others turned out to be the real deal: vast collections of galaxies and dark matter in a great gravitational equilibr
ium.

  As time went by we pushed ever farther out into the cosmos, finding these close-knit galactic communities at greater and greater distances. The ultimate prize we sought was to use these objects as a surrogate set of scales for weighing the whole universe: we literally wanted to measure the mass of the cosmos. It was an ambition that many astronomers had long pursued, and in essence the idea is simple. The more mass the universe contains, the more quickly galaxy clusters should appear to grow. Most of that growth should also take place more recently, within the past few billion years, if the universe is chock-full of matter. This means that in a weighty universe we might expect to find very few, if any, clusters, at great cosmological distances. Conversely, in a universe that contains less matter, our measures of cosmic distance and time are different, and the growth of galaxy clusters appears as a weaker and more prolonged affair. By finding lots and lots of clusters, we aimed to refine the statistics and to narrow down estimates of the total content of normal and dark matter in the universe. In doing so, we would arrive at a comparison between science’s most fundamental cosmological models and nature itself.

  For all of us it became a bit of an obsession. Ebeling, Jones, Perlman, and I seldom went a day without dealing with some piece of the puzzle. After a time we were also joined by Donald Horner, a hardworking graduate student from the University of Maryland. Together we’d pore over the output of the computer algorithms that sniffed through the mountains of X-ray imagery. We’d argue about what looked real and what didn’t, flinging printouts of visible-light pictures of what might be galaxy clusters at each other. Then periodically we’d go off to big telescopes to finally nail down our very best candidates, sitting bleary-eyed through long mountain nights in Arizona, Chile, and Hawaii. It wasn’t a process that wrapped up quickly, especially since pushing our observations to find ever more distant systems was paramount. There is enormous leverage in distinguishing between different cosmological models if you can see how rapidly clusters form in the young universe. Finding even a single cluster at earlier and earlier cosmic times can make or break certain theoretical scenarios. It’s like finding the fossil remains of a feathered dinosaur millions of years before you expect to. Eventually these rare and unique cases might force our ideas about evolution to change.

  But galaxy clusters have some tricky characteristics. While I might describe them as “objects” in space, the truth is that they’re not like planets or stars. Those bodies are highly self-contained and have a quite well-defined moment at which they finish their formation. A star is a star when its fusion engines are fully started. A planet is a planet when it ceases to accumulate a noticeable amount of material. But a galaxy cluster is a great amoeba of gas and stars, hanging at the intersections of a larger webbing of structure that extends for hundreds of millions of light-years. Matter accumulates almost continually from the surrounding universe, heating as it falls faster and faster inward. Optimistically, we might consider a cluster complete once its constituents reach a state of physical equilibrium. Hot gas will sit quietly in the deepest regions of its gravitational well once the forces of pressure and gravity balance out, and galaxies will remain within the system once their orbits are established. But we know that nature can be awfully messy. The gas cools; black holes throw out energy; incoming matter slowly adds to the system. The precise moment at which a cluster “becomes” itself is therefore open to some interpretation. In astronomy, a field so utterly dependent on observation, nothing beats going and looking. So to answer questions about the baby steps of galaxy clusters, the very best thing would be to find these infants in the act of growing up. And this was where serendipity would rear its head.

  Late in the summer of 1999, I made the long journey from the United States to an astronomy conference on the volcanically hewn Greek island of Santorini. It would be tedious to go into how great a place it was for a bunch of sunlight-deprived scientists to hang out, but it was a special thing to be in such dramatic surroundings to discuss the latest science. Also attending was an old colleague, the English astronomer Ian Smail from Durham University in northern England. Smail had made his name chasing some of the most distant, and hence youthful, objects in the universe. In the latest advance, he and his collaborators were making use of an intriguing and rather new type of astronomical camera. Unlike typical devices for making images of the sky, this camera operated in a netherworld of the electromagnetic spectrum. Between infrared wavelengths and the beginnings of microwaves is a spectral region known as the submillimeter. It’s a notoriously tricky regime. At slightly higher energies and shorter wavelengths we can treat light as bouncy photons, trapping and focusing them with our mirrored telescopes. At slightly lower energies and longer wavelengths, we have to treat it as waves that require antennae for detection. Lurking in between is the realm of the submillimeter. It’s a slippery beast. (Electromagnetic radiation in this range can penetrate through a person’s clothing before reflecting off the outer layers of our skin. Because of this, it is a critical component of some of the machines used to scan your body for concealed items when you go through an airport screening. The stealthy submillimeter photons allow a discreet amount of electromagnetic frisking.)

  For astronomy, it’s also a regime where our Earth’s moist atmosphere is full of potentially confounding noise from the wiggling energetics of water molecules. Submillimeter radiation from cosmic sources is mostly absorbed and overwhelmed by this curtain-like barrier. There are only certain spectral “windows” that are clear enough to look through, and less than half a dozen places on Earth where the environment is dry and dark enough for us to have a hope of peering into the submillimeter universe.

  Nonetheless, new technology had produced a camera that could now make astronomical images in this tricky spectral range. The particular device that Smail and his colleagues were using was on the 2.5-mile-high peak of Mauna Kea in Hawaii, attached to the appropriately named James Clerk Maxwell Telescope. It was well known that out in the nearby universe, cold rich dust and gas was an excellent source of submillimeter radiation. The thickly blanketed and mysterious environments of star-and planet-forming nebulae, or dust-rich nearby galaxies, were perfect targets. But Smail and his group were interested in places far, far removed from these.

  He and his group wanted to push back into the deep history of the universe, and the submillimeter spectrum offers a unique vantage point. As light traverses the cosmos, its wavelength gets stretched. Our universe is expanding, and spacetime is inexorably swelling. Galaxies far enough apart to have minimal gravitational interaction are flying ever farther away from one another, like raisins in an endless sea of rising yeasty dough. The very tissue of the universe that light must travel through is widening, and a photon that was once at an ultraviolet wavelength can arrive on the other side of the cosmos as one of visible light. A photon that began its journey 10 billion years ago as a blip of infrared energy will arrive as a little wave of submillimeter radiation. Incredibly, Smail’s collaborators were finding faint mounds of this cool radiation on the sky that consisted of the cosmically stretched photons of infrared light coming from the rich, dusty, gaseous mix at the very birthplaces of galaxies and stars. These were objects, perhaps protogalactic structures, whose light had taken more than 10 billion years to reach us. As we sat basking in the eight-minute-old photons of brilliant Greek sunshine, we talked excitedly about their most extraordinary finds.

  They had taken images of locations that were already known to be strong sources of radio waves in the youthful universe—precisely the kind of objects associated with supermassive black holes in our own more mature cosmic neighborhood. In these places they were discovering great regions of submillimeter emission, tens of thousands of light-years across. These were the hallmarks of clouds of dust that were tens of millions of times the mass of the Sun. They were heated by the intense radiation of newly forming stars, and perhaps something else, shrouded inside. Some of these dusty places were also grouped and clustered together. T
he statistics used to come to this conclusion were a bit threadbare, but solid enough to take a chance on. It looked like they could be the toddlers that were going to grow into the overfed hulks of galaxy clusters found in our modern universe.

  Smail and his colleagues were keen to determine whether or not there really were supermassive black holes lurking in and around their distant sightings of warm material. Such behemoths would also help heat up the masses of dust that were producing the submillimeter light, and would be critical ingredients to understand. I was keen to find out whether these really were the locations of youthful galaxy clusters, the beginnings of those great cathedral-like structures of stars, gas, and dark matter. If they were, then perhaps we could learn something of those structures’ initial building blocks.

  There was an obvious way to pursue all these goals, and that was through an X-ray telescope. Only the most energetic and penetrating X-ray photons stood a chance of drilling out through the dusty cloak surrounding a massive black hole in such a place. And the fog-like X-ray emission of hot gas getting trapped inside a growing cluster’s gravitational bowl was also the best way to measure that great warp in spacetime. We clearly needed to look for X-rays from one of these submillimeter mysteries, and luckily our timing was good. Chandra, NASA’s high-performance, $2 billion X-ray space telescope, had just been launched a few months before I met with Smail in Santorini. It was the ideal instrument to chase these distant objects. We just needed to choose the best target. It wasn’t a hard choice to make: it had to be the brightest of the distant and dusty blobs, with the uninspiring name of 4C41.17. This mysterious form was a mind-boggling 12 billion light-years away.