Cosmic Dawn: Astronomy's Ancient Quest to Find the Universe's First Stars

"I only care about the red ones," he said. "Everything else is foreground."


The Hubble Ultra Deep Field, courtesy of the Space Telescope Science Institute.

One morning this fall, I found myself drinking tea with Massimo Stiavelli, the Project Scientist for the James Webb Space Telescope, staring at one of the most famous and profound images produced by the Hubble, the Ultra Deep Field. In his office at the Space Telescope Science Institute in Baltimore, Stiavelli has a large, glossy print of this penetrating gaze into the most distant recesses of the observable universe.

Which makes sense, because in 2004, his team created it.

They pointed the Hubble at a tiny corner of the sky, letting it pile up light for ten days. The resulting image contains more than five thousand shimmering galaxies, many of which are beautifully detailed. I began asking Stiavelli about some of the larger ones, the gaudy cosmic baubles clustered here and there throughout the image. But he said he had no use for them.

Instead, he pointed to a tiny red dot near the top of the image, a dot so small you can't see it in the low-resolution version above.

"I only care about the red ones," he said. "Everything else is foreground."
Stiavelli is one of the world's foremost experts in first light, a subfield of astrophysics devoted to studying the first stars that formed after the Big Bang. He likes tiny, red galaxies because they tend to be old and distant. Indeed, the more distant a light source is from Earth, the more stretched and red its light appears to us -- a phenomenon we call redshift. Some light is so distant that it stretches out of our visible range and into the infrared, where only a specialized telescope can see it.

The redshifted light of Stiavelli's little dot tells us that it left its home galaxy when the Universe was less than a billion years old. By the time it alighted upon the Hubble's mirror, it had been traveling for more than twelve billion years. A long time to be sure, but not long enough to qualify as first light. The stars that made this dot had ancestors.

One way astronomers determine the relative age of a galaxy's stars is by using spectroscopic analysis. The trick is to run an ancient galaxy's light through a spectrometer to test its metallicity, because metallicity is an excellent proxy for star turnover. Stars are powered by thermonuclear fusion, a process that generates new elements, including metals. When they die, they explode the rich, metallic contents of their cores into the surrounding cosmos. The more stars that explode in a galaxy, the more metallic its makeup. First light hunters like Stiavelli want to find "pristine galaxies," early collections of pure hydrogen-and-helium stars, untainted by the shrapnel of their exploded ancestors.

Today those stars -- the suns of cosmic dawn, made rosy-fingered by redshift -- lay just beyond the reach of our scientific instruments. But, only just. Before the decade is out, we expect our telescopes to be able to see them. By that time, our technological eyes will have adjusted to the darkness of the deepest corners of the Universe.


Ours is not the first generation to wonder about light's origins. Intellectual interest in first light preceded both modern astronomy and the invention of the telescope. Western philosophers like Plato and Aristotle thought that stars were timeless, that they had always existed as fixtures of an eternal cosmos. Buddhist and Hindu cosmologies both describe a regenerative Universe, whose stars are subject to infinite cycles of creation and destruction. Medieval Christian theologians like St. Thomas Aquinas were fascinated by the creation of light by fiat, as described by Genesis 1:3 -- "And God said 'let there be light,' and there was light." In the 4th century, an Archbishop from Milan named St. Ambrose went a step further, by daring to put forth a teleology of light, a reason for its existence. He said that God created light to "reveal the world by infusing brightness therein . . . to make its aspects beautiful." Modern science has done much to satisfy this ancient curiosity about first light. Indeed, cosmologists are confident they can now describe the precise physical conditions that gave rise to it.

After the initial explosion of the Big Bang, the Universe expanded and cooled for hundreds of millions of years, a period we call the cosmic dark ages. With time, parts of that sea began to aggregate, forming regions of condensed gas more massive than a hundred billion suns. Within these regions -- called haloes -- clouds of hydrogen contracted into stellar seeds that grew denser as they slowly surrendered to gravity. Once the stellar seeds reached a certain density, their cores erupted into fusion reactors. Soon after, fresh starlight began pouring out from their surfaces, illuminating the young Universe.

As The Atlantic's Megan Garber explained earlier this month, much of that early starlight still exists today, dispersed across deep space like cosmic fog between galactic streetlights. But what if you rewound all that light, returning it to the primordial stars that produced it---what would those stars look like?

"Though we don't know exactly what these first stars looked like, there is evidence that they would be very massive," Stiavelli told me. Some cosmologists think they were as large as thirty suns put together; others, as many as five hundred. The uncertainty arises from our inability to adequately model the complex processes of early star formation. We're confident about the structure of the large hydrogen haloes that produced the first stars, but we're not sure how they fragmented over time. We're not sure whether the haloes produced a few super-suns, or a multitude of smaller objects.

"We've recently advanced to the point where we can model a thousand years from the moment the stellar seed starts accreting," Stiavelli told me. "But a thousand years is a short time for a star; to have a complete picture, you want to see hundreds of thousands of years."

Despite these limitations, cosmologists suspect the first stars burned hot and, as a consequence, were exceptionally short-lived.

"The effective temperature of the sun is just below 6,000 degrees Kelvin, but the effective temperature of the first stars would be around 100,000 Kelvin," Stiavelli said. "That's hotter than any star in our galaxy." Astrophysicists predict that these scorchers would have burned themselves out in less than 3 million years -- a blip compared to main sequence stars like our sun. These hard-living stars quickly seeded galaxies with heavy elements, the building blocks of rocky planets and carbon-based life, but their short life spans make them elusive targets for first light astronomers. After all, how do you see a star that shined for less than a twinkle of cosmic time?

Alas, help is on the way. In 2018, the James Webb Space Telescope is scheduled to launch. The Webb is a massive, space-based infrared observatory, an unprecedented savant in the search for stretched light. Its infrared sensors will be sensitive enough to measure the chemical properties of every object in the Ultra Deep Field -- even the red dots.

"Our knowledge of this deep core of the Universe is about to get extremely detailed," Stiavelli told me. The Webb will also see more ancient objects, galaxies too old and faint to be picked up by the Hubble. Its Ultra Deep Field could be freckled with red dots, and may well reach back to the cosmic dark ages.

If not, there is another way first generation stars could catch the Webb's eye: by exploding. We don't yet know how many first generation stars exploded into supernovae. But we do know the ones that do are likely to be quite luminous, and we expect their brightness to linger for years. If the Webb glimpses one of these objects just after it detonates, it could observe it over time, watching to see the extent of its flash, and the way its light interacts with the cosmos around it. Astronomers could follow it for months, or even years, until it finally dims, beginning its slow dispersal into cosmic fog.

But the Webb is still half a decade away.

In the meantime, cosmologists working with existing infrared telescopes have devised another ingenious way to hunt for first light, a work-around that leverages the exotic, light-bending properties of gravity. On a large scale, our universe is organized into clusters -- collections of galaxies so gigantic that their collective gravity can divert light itself, bending and stretching it like a lens. Point a telescope at one of these clusters and you can catch light from its far side, light that gets magnified in the process of being wrenched around the cluster. In a way, these massive objects are like organically occurring telescopes, natural pockets of magnification embedded throughout the cosmos. The only catch is that you can't always predict what these visual wormholes will magnify.

"In any lens, there are points where the amplification is a thousand or even higher," Stiavelli told me. "But the probability that one of these individual stars will correspond to these points is very small."

Among first light hunters, the game is to be the first to identify a target for the Webb, the first to slap a specimen onto a slide, so that Webb can zoom in and see its guts. Some astronomers are trying to improve their chances by pushing existing infrared instruments to their absolute limits. Already, we've laid eyes on twelve-billion-year-old supernovae. It's not unreasonable to think we'll see something older before the decade is out.

Stiavelli is hoping we get lucky.

"The ideal situation would be to find a primordial supernova in 2016 or 2017, so that it's still glowing when the Webb goes up in 2018," he said. "That way we can follow it up and get spectroscopic information, so we can be certain it's a first generation star and not something else." 

The Webb was designed to be a first light observatory like none that have come before it. One way or another, it's going to set its sights on cosmic dawn, on the earliest sunrise we can yet imagine. But it would be nice to get a head start or, at least, to know where to look.


Artist's concept of the first stars winking alive in the gloom after the Big Bang. Courtesy of NASA/JPL-Caltech/R. Hurt (SSC)