Physics & Math / Cosmology

A globular cluster floating in the galactic halo of the Milky Way, only 7,000 light-years from Earth, is home to some of the oldest stars ever seen with an optical telescope. Globular clusters, which can contain up to 1 million densely packed low-mass stars, are the progeny of the universe's first stars, huge hydrogen fireballs that flamed out as supernovas. For a close-up of this cluster, see below. Photograph courtesy of NOAO/AURA/NSF.

Fourteen billion years ago, give or take an aeon, the Big Bang created the universe—or did it? If by universe you mean an abysmally black void with no stars, galaxies, planets, or the slightest promise of life, then the Big Bang's your baby. But if you mean the starry cosmos we see around us today, a universe that includes at least one planet with life, the Big Bang was a bust. For 100 million years after time began, the universe was less interesting than a mud puddle. Only a handful of elements—mostlyhydrogen and helium, along with faint traces of lithium and beryllium—ricocheted through fathomless, unending gloom. Had anyone been around at the time to bet on the future, the smart money would have been on more of the same: darkness, emptiness, death. Yet improbably, miraculously, the universe—the ultimate dark horse—beat those odds. It was reborn. A 100-million-year-long night ended when clouds of hydrogen collapsed and ignited. In the blast furnaces of the first stars, atoms were crushed, burned, and transmuted into more complex particles, like the carbon in the paper of this page or in the hand that's holding it. That moment—when the universe first lit up—was nothing less than a second creation, the one that really counts.
    Astronomers, however, haven't yet been able to see that cosmic dawn, because it's well beyond the range of any existing telescope. "Ah, but a man's reach should exceed his grasp," wrote Robert Browning, "or what's a heaven for?" Or, as astrophysicist Tom Abel might say, "What's a computer for?" Abel has not extended our grasp to heaven, but he has reached out to the time and place when light transfigured a murky universe. He and two colleagues did it with software, using an astonishing computer program to resurrect a long-ago epoch dominated by giant balls of flaming hydrogen hundreds of times bigger than the sun and millions of times brighter. They were the first stars, unlike any in the universe today. They created everything necessary for all future stars, as well as the essential elements of life as we know it on Earth. These fireballs blazed for about 3 million years and then died in a chorus of detonations, long before anything we now see in the sky existed. And their death created life. Parts of those earliest giant stars are in our blood, bones, and skin. It may even be that a bit of the stardust of which the Earth is made was shot into the void by the explosions that ripped apart the earliest stars.

In a nondescript concrete building on the campus of Pennsylvania State University, a six-foot-tall matrix of animate stardust named Tom Abel is looking much farther out into the universe than any telescope ever has. Abel is 32, born and educated in Germany. His English is fluent. He wears an untucked, loose-fitting white shirt, black jeans, sandals, and has a gold earring in his left ear. When he's not re-creating a universe, he likes to sky-dive—his Web site features a photo of him plunging from a plane; he appears to be screaming.
    Now he's on solid ground, observing a cosmos on the flat-panel display of his desktop computer. He has a god's-eye view of the universe 100 million years after the Big Bang, when it was just a thirtieth of its current size, and when a cosmic dark age was about to end. Multicolored eddies of hydrogen gas fill the monitor, like psychedelic cigarette smoke, color-coded to show different densities. Tens of thousands of years flash by in a few seconds as the churning gas clouds begin to coalesce into a nascent fireball. At the moment, the monitor encloses a virtual space-time (see "What is Space-Time?," below) 20,000 light-years across, about one-fifth the diameter of our own galaxy. But Abel can zoom in on smaller domains in his creation.
    The ability to shift focus effortlessly over enormous scales of space and time is the simulation's key feature. Every image that appears on the computer monitor is built from a grid consisting of thousands of individual cells. Within each cell, the computer solves dozens of equations involving gravity, heat flow, and collisions between atoms and molecules. When Abel chooses to zoom in on some region, even if it's only a thousandth of the size of his previous view, the resolution of the image doesn't change or become fuzzy like a magnified photo. The underlying grid still consists of thousands of cells, with the physics meticulously worked out in each one. The simulation is like an extraordinary microscope that never goes out of focus, no matter how large or small the target object. What drives this impressive programming is not computing power but a detailed understanding of stellar physics from the level of molecules all the way up to gravitational interactions spanning light-years.
    "Each time we had to stop our research, it wasn't because we didn't have a big enough computer," says Abel. "It was because we ran out of physics. Greg Bryan, who's now at Oxford, wrote the program that allows us to focus in on regions as they collapse. We don't lose any resolution. If the largest volume of our computer simulations contained the entire Earth, our smallest grid would be the size of a red blood cell in your body."
    Abel, Bryan, and physicist Michael Norman at the University of California at San Diego have been working on this desktop universe for seven years. Their work has not only transformed astronomers' views of how the universe first lit up, it also marks a major change in how astronomers do their work.
    Until recently, cosmology was primarily driven by observations—an astronomer detected a mysterious phenomenon and then tried to figure out what it was. In the late 1950s, for example, astronomers discovered powerful sources of radio waves all across the sky. They were labeled quasars, for quasi-stellar radio sources. But years passed before anyone understood that they are the bright cores of very young galaxies.
    The work of Abel, Norman, and Bryan reverses that process, which happens to make cosmology more predictive. The universe, when modeled on a computer, becomes a laboratory where astronomers can test their theories. By creating models of stars that lie beyond the range of today's telescopes, Abel's group can tell astronomers what they might find 10 years from now. So what will they see when telescopes are better?
    Abel shifts in his seat and starts the simulation with a click of his mouse. It begins like this: The monitor shows a dark cosmos—the universe as it was 13 million years after the Big Bang. Like an ocean surface, the darkness conceals powerful currents. Something stirs in the depths. Dark matter, which had been randomly scattered throughout the cosmos, starts to clump together. Astronomers estimate that this mysterious material accounts for more than 90 percent of the matter of the universe. They know it's out there only because they can detect its effect on the motions of galaxies. But they can't see it directly, and they don't know what it's made of.
    The screen fast-forwards 90 million years. It's still dark, but now the simulation shows a universe with structure. Strands of dark matter stretch across the cosmos like a giant web. And there's another component, almost unnoticeable, floating like the thinnest of mists through space: a gas, mostly hydrogen. Pulled by gravity, the gas begins to form clouds around the densest regions of dark matter. Each cloud is stocked with enough matter to make 100,000 suns. By comparison, our Milky Way galaxy contains a few hundred billion suns.
    Abel lets another 50 million years slide by, and then zooms in on one of the hydrogen clouds, narrowing his field of view from 18,000 light-years to 1,800 light-years across. If we were looking through the porthole of a rocket ship instead of at a computer monitor, we would have to travel faster than the speed of light for the view to change so rapidly. The cloud itself fills a third of the screen; it's about 600 light-years wide and still pregnant with enough hydrogen to make 100,000 suns. And now, for the first time since the fires of creation cooled after the Big Bang, warmth enters the universe: The entire cloud has started to glow. In the simulation, it looks almost like a flower, an orange poppy, fragile and small against the black velvet of dark matter and empty space. The glow is from gases compressed and heated by gravity as they fall toward the cloud's center, where the first star will be born. The heating is simple backyard physics—compressed gases get hot, like air pumped into a bicycle tire. But it's backyard physics on a titanic scale: Every hundred years, enough gas to make Earth's sun plunges into this embryo.

    Most of the cloud is comfortably tepid, about room temperature—albeit in this case a room so large that light needs six centuries to cross it. That amniotic warmth is far different from the frigid -440-degree clouds where stars hatch today. Molecular hydrogen is a terrible coolant. Once it warms up, it stays warm and loses heat very slowly. This has a fateful effect on star formation. Before a star's nuclear bonfires can ignite, the stellar embryo—the core of the star's parent cloud—first has to collapse into a ball hot and dense enough to fuse hydrogen atoms. However, like all warm things, the hydrogen in the cloud tries to expand, fighting the gravitational pull necessary for the star to burst into life. For gravity to overcome this internal pressure, a critical mass of several hundred suns must build up in the cloud's core. In Abel's model, with roughly the equivalent of a sun per century raining onto the core, the process takes about 100,000 years—a humanly imaginable time, about half as long as our species has been around.
    With a thousand suns' worth of hydrogen on the scene, you might wonder how many stars finally emerge from the mother cloud. It's a question many astronomers have asked. Until Abel, Bryan, and Norman came along, no one had a solid answer.
    The trouble with earlier predictions was that none of them could account for all the complex steps involved in the transformation of a cloud hundreds of light-years in size to something as relatively puny as a star. At some point, calculations simply gave way to educated guesses. Abel, Norman, and Bryan took away the guesswork. But even they couldn't predict the outcome of the simulation before running it for the first time.
    Abel turns back to the monitor and this time zooms in closer to the core of the cloud, first to 180 light-years across, then down to 18. Finally, we're looking at a region a mere six light-days wide, about 20 times the size of our solar system. At each step, the core remains intact; it doesn't break up. The cloud, which could have made a thousand stars, sires just a single newborn, but it's a whopper, roughly 200 times bigger than our sun.
    In hindsight, says Abel, the result seems obvious. The pressure of expanding hydrogen in the cloud slows the gravitational collapse, making it smooth and uniform. That's very different from the way stars are born now, in clouds filled with a stew of elements created in the blast furnaces of earlier stellar generations. Besides hydrogen, there's usually iron, carbon, and oxygen; those elements all dissipate heat much more efficiently than hydrogen. Clouds that contain them cool and crumple more readily than clouds of hydrogen. Because these elements are scattered unevenly throughout the cloud, different parts collapse at different rates, creating clusters of small stars instead of the lone behemoths that first illumined the cosmos.
    Even with billions of these behemoths ablaze, the universe remained immersed in darkness. A light-absorbing hydrogen fog—leftover gas that never condensed into star-forming clouds—filled all of space, muting the primordial starlight. But once they got going, each of the early stars poured out a million times as much radiation as the sun, and that began to burn the fog away. Their radiation ripped the hydrogen atoms apart, splitting each into a single electron and a naked proton. Gradually, the burning balls of gas carved out ever-wider bubbles of luminosity. It probably took a few hundred million years—and another generation of stars—for the hydrogen to dissipate completely.
    Abel hasn't yet simulated the death throes of the first stars, but early calculations indicate they would have burned through their nuclear fuel a few thousand times more rapidly than our sun. They were cosmic SUVs compared with our fuel-efficient hybrid, the sun. They lived fast and died young. After only 3 million years—a life span so brief it doesn't match the length of time since the appearance of the earliest hominids—the stars would have exploded as supernovas. And this set in motion the evolution of the universe as we know it.
    The catastrophic explosions scattered carbon, oxygen, and iron—created in the stars' cores by a chain of fusion reactions—across space. These elements, which trigger the collapse of star-forming clouds, were the seeds for all future stars and solar systems. Without the utter destruction of the universe's first lights, life never would have arisen on a small, watery world billions of years later. None of us would be here to wonder about how the stars were born if they hadn't died.
    "Now we know that these first stars would become supernovas," says Abel. "We didn't know that before. It's exciting because we can try to find them observationally."
    Abel is especially intrigued by the tantalizing possibility that astronomers may even now be witnessing the long-ago demise of the first stars. For several decades, powerful flashes of gamma rays—high-energy radiation—erupting at the very edge of the visible universe have puzzled astronomers. They are the most powerful phenomena ever observed, pouring out in a few hundred seconds as much energy as a typical supernova does in months. One of the leading explanations for the gamma-ray bursts is that they are produced during a supernova explosion, when the core of the detonating star caves in on itself so violently that it creates a black hole. Material from the star that would otherwise have been blown into space by the supernova instead cascades into the black hole and is heated to such temperatures during the fatal plunge that it gives off gamma rays. If some of the first stars did die this way, their corpses might still litter the cosmos as black holes, some of them now swollen to enormous proportions after aeons of swallowing dust and even entire stars. The black holes left behind by the first stars may have even become the gravitational anchors around which early galaxies coalesced.
    Niel Brandt, who works just a few doors down from Abel, thinks he may have some pictures of those black holes, or at least the brilliant young galaxies called quasars that harbor them. Brandt picks up several photos taken by Chandra, a NASA X-ray satellite that has been orbiting Earth for 31/2 years. "These are the three most distant black holes known to date," he says, pointing at some small white blobs on photos filled with many white blobs on a black background. The photos capture not the black holes themselves but the X rays emitted by the doomed gases and stars falling into the holes in the centers of the quasars, the farthest of which lies 13 billion light-years from Earth.
    "The very massive stars that Tom studies are probably some of the objects that collapsed to form the seeds for the black holes that we're seeing with Chandra," adds Brandt. For his part, Brandt seems astonished by what he has been able to learn so far from Chandra, given the scanty nature of the data. The photos he holds were created by photons—particles of light—that barely escaped entombment in a black hole. A journey that began at least 13 billion years ago—nearly three of Earth's lifetimes—ended when the photons hit an inch-long electronic detector aboard Chandra. "We're talking here about a total of 6 to 20 photons reaching the detector," says Brandt. "Basically you get a photon coming in every 15 minutes or so, but Chandra is so sensitive, it can still detect them."
    The disparity between the images, tiny splotches of 20 photons at most, and the violent reality of the distant black holes, objects capable of shredding entire stars, beggars comprehension. It's like trying to imagine the power of a tidal wave from the feel of a gentle ocean mist on your face.
    The Chandra satellite found those three black holes in a small region of the sky in the constellation Ursa Major, first mapped by the Sloan Digital Sky Survey. Brandt picks up a copy of a Hubble photo to show the area Chandra surveyed.
    "This is quite an amazing picture," he says. "The portion of the sky contained in this image is only about as big as if you took a dime and held it at arm's length and looked at Roosevelt's eye. It's the size of that eye held at arm's length. A very tiny region. Yet in that region you have something like 3,000 galaxies. We've now found tons of X-ray sources there. And by doing further surveys of this type, we can hopefully find very distant black holes, black holes closely related to the type of things that Tom studies."

Dim white dwarf stars 12 billion to 13 billion years old are sprinkled amid brighter, younger lights in a Hubble close-up of a globular cluster. A typical globular cluster star contains more heavy elements than any of the hydrogen fireballs that first lit the universe but about one ten-thousandth the amount of iron found in later-generation stars like our sun. Photograph courtesy of NASA and H. Richer/University of British Columbia.

Five time zones away from the Penn State campus, on the big island of Hawaii, a friend of Tom Abel's is having a frustrating night. Harvard astronomer Kurt Adelberger is waiting to see if the weather will clear. He's visiting the small town of Waimea, headquarters for the Keck telescope, the world's largest. High-speed data cables, like long optic nerves, snake to the summit of Mauna Kea, 11,000 feet above Waimea, linking Adelberger's computer to the Keck's great eye, a 33-foot-wide mirror. Like most astronomers who use the Keck, Adelberger doesn't work at the summit—the thin air is great for observing the distant universe but terrible for thinking. The headquarters for the world's preeminent telescope sits on the edge of a small shopping mall, 50 yards from a McDonald's.
    For the past two years, Adelberger has been studying about a dozen distant galaxies from a time when the universe was 2 billion years old, just a stellar generation or two removed from the first stars. He has found evidence that powerful winds, spawned by supernovas, racked these galaxies, blowing stellar material as far as half a million light-years from the galaxy. The winds, created by the explosions of perhaps 100 million supernovas in each galaxy, may have lasted for 100 million years. "In our galaxy, we get about one supernova every century," says Adelberger. "But if you had lived at the time we're looking at, you would have seen one a year. So you're getting a hundred times more. Stars were forming much more rapidly in the past. Most of the stars that are ever going to exist formed long ago, in the first half of the universe's history.
    "One of the things I think is interesting is we want to know where the elements we're made of come from, because all the Big Bang made was mostly hydrogen and helium. But if you look around the room, there's not much of that. Everything's mostly carbon or iron, and where does that stuff come from? Well, the only place it's made as far as we know is in stars, in these big stars that blow up. So everything in this room must at some point have been part of a big star that then blew up and got scattered into the universe. You might ask, Well, how far away was this star that blew up that we were once part of? I think people would have thought until recently that this star was probably somewhere in our own galaxy, not all that far away. But the kind of thing we're seeing says instead that the star we came from might have been in another galaxy altogether, and the supernova winds crossed through space and sent materials to different galaxies."
    Adelberger is in Hawaii to make more detailed observations of the supernova winds. But competition for time on the Keck is keen, and he has been granted only five nights of observing time. Four of those nights have already passed. Tonight, hail, rain, and clouds shroud the 13,796-foot summit, blocking light that has traveled for more than 10 billion years. Adelberger won't get to see his galaxy, not on this watch. Tonight the eight-story-tall, 300-ton telescope, eternity's looking glass, reflects only the dark walls of the shuttered observatory. Outside, above, and around Keck's sheltering dome, not a single star can be seen. On nights like this, Adelberger envies Abel.
    Abel, meanwhile, has more plans for his telescope-free practice of astronomy. He would like to create a virtual telescope modeled on the James Webb Space Telescope, the successor to the orbiting Hubble telescope. Formerly dubbed the Next Generation Space Telescope, the JWST isn't scheduled for launch until 2010, but with the right computer program, Abel thinks it might be possible to predict what that telescope will see years before it's off the ground. "We're hoping to get funding to start building virtual telescopes—we would do observations of a virtual computer universe in such a way that it would perfectly mimic what the JWST may do."
    With any luck, the JWST just may help Kurt Adelberger and Niel Brandt understand how a universe filled with giant black holes, young galaxies, and supernova winds emerged from the one so realistically re-created by Abel.
    "The challenge is to understand how, in just a billion years, between what Tom studies and what we can actually see, you actually manage to do all that," says Brandt. "A lot happened during that brief epoch in the very early history of the universe, and untangling what happened then will take probably the next decade or two to really pin down."



Discover 101: What is space-time?

In our everyday lives, space and time seem to be two very different things: Where is not the same as when. But in 1908, Hermann Minkowski, a German mathematician and one of Einstein's teachers, proposed that space and time form an indivisible whole.
    We live, Minkowski said, in a four-dimensional world: three of space and one of time. When we see the sun, for example, we're looking not just at an object in space but at a time as well. We see the sun not as it is but as it was eight minutes ago, the time it takes the sun's light to reach Earth. Space does not exist separately from time. Or as Minkowski put it, "Nobody has ever noticed a place except at a time, or a time except at a place." Astronomers now take for granted the union of where and when. They measure distance in light-years, and when they gaze into the depths of space at the most remote galaxies, they are simultaneously peering back in time. In a sense, they're looking at a ghost universe because the light hitting their telescopes has been traveling for billions of years, and it came from stars that may no longer exist. Unlike archaeologists, astronomers can watch the past unfold before their eyes.
— T. F.




Tom Abel, Greg Bryan, and Michael Norman's research on the first stars is mapped out with images, movies, and news articles at www.astro.psu.edu/users/tabel/GB/gb.html.

For a slightly technical primer on first-star research, go to www.arcetri.astro.it/science/cosmology/RTNfirststars.html.