God In The Equation

by Corey S. Powell

  While Perlmutter was painting a picture of the runaway universe to an enthusiastic press corps, the High-Z Supernova Search was keeping a low profile. The group did not yet have a completed paper making its own case for cosmic acceleration. When Schmidt's results went public at a cosmology conference a month later, most of the media treated it as confirmation of a story they had already reported. Perlmutter casually reinforces that view: “Our competing group came in with a result just a few months later that agreed well with ours.” Schmidt bristles at being seen as an also-ran and focuses on when the groups were doing their research rather than when they announced their results. “To be fair, we came out with the answer at the same time. We ended up converging on the same answer, though their view has ended up shifting over time,” he says. Kirshner sounds less diplomatic, accusing Perlmutter of misleading people about the chronology of events: “For some reason, he's eager to establish it wasn't a dead heat, but it was.” At the end of 1998, Science magazine recognized both teams when it named the discovery of the accelerating universe its “breakthrough of the year.”

  Given the intense competition between the Supernova Cosmology Project and the High-Z Supernova Search, astronomers were particularly impressed that they had arrived independently at almost the exact same result. This is one of the key ways in which sci/religion trumps the old-time religions. It airs its differences openly to help sort out the biases, hopes, and dumb mistakes that influence its discoveries. In its incremental and sometimes blundering way, sci/religion follows the light that leads toward ultimate, unobtainable truth.

  The two teams developed distinctive ways to interpret their galactic images, evaluate the light curves, and translate them into distances; they also mostly worked with different data, although they did share two supernovas in their early results. Yet they came up with almost identical statistics on the density of the universe. According to the supernovas, all the matter in the universe adds up to one-quarter to one-third of the critical density, the amount of matter necessary to halt the motion of the big bang. That in itself is a significant finding, because it means that the expansion will continue unchecked forever unless the physical state of the universe changes drastically. The new results largely eliminate the hope of a “big crunch” that will start a new cycle of existence. The supernova data also jibe well with the increasingly refined methods for weighing the universe, such as counting the numbers of galaxy clusters and measuring hot gas around the clusters held in place by their gravity. These techniques showed that all the matter, visible and dark, comes to about one-third of the critical amount, closely matching the supernova's story.

  The true stunner was that Perlmutter and Schmidt also agreed that we live in a runaway universe where Lambda, not matter, controls our destiny. “That seemed to me pretty shocking,” Gum says. Nobody from Einstein on down had ever seriously proposed such a possibility. Einstein surely would have been baffled by the return of his greatest blunder, which he had created to bring order to the universe and abandoned to bring back common sense. Yet this discovery continued Einstein's mystical quest by showing that space is an equal partner with matter, and both are accessible to human investigation. Although Lambda is energy, not matter, it has an equivalent mass and density (another consequence of Einstein's E=mc2). The Supernova Cosmology Project and the High-Z Supernova Search processed their numbers and found that their results fit very nicely with a universe in which one-third of the density is matter and two-thirds is an unknown form of energy, possibly Lambda. The total density would then be exactly the critical value, just as Gamow and Linde had proposed two decades earlier in the inflationary theory of the big bang.

  The amount of energy contained within the hypothetical Lambda is overwhelming on cosmic scales but minuscule by human standards. If you scooped up a block of empty space 250,000 miles on a side—about the distance from the earth to the moon—you'd find just about one pound of energy inside, assuming you could find a magical technique for weighing it. In that same box of space, you'd find roughly half a pound of ordinary matter, mostly hydrogen atoms. The universe is very nearly empty. The battle between Lambda and matter is thus a fight between almost nothing and even closer to nothing. But the conflict has an important implication for the history of the universe. If the universe is accelerating now, its expansion was slower in the past and its age is greater than one would expect simply by extrapolating backward from the present motions. With Lambda in the equation, Perlmutter places the cosmic age at fourteen billion to fifteen billion years, a nice, comfortable range that allows plenty of time for the first stars to form and that settles the “age paradox” that has bugged cosmologists ever since Hubble.

  Theorists immediately went to work trying to understand this new development. Vacuum energy could come in many forms. Michael Turner of the University of Chicago, a thoughtful and outspoken cosmologist, coined the generic term “dark energy” because, he says,“ 'funny energy' didn't sound serious enough.” The accelerating universe demanded respect. At the American Astronomical Society meeting, Perlmutter was still stunned by significance of the supernovas. “Even as we were talking about it, we hadn't realized the full enormity of what we were saying,” he explains. Schmidt was similarly befuddled. And Kirshner recalls talking about cosmic acceleration with his former students in the High-Z team: “I thought it was a terrible idea. I didn't think it was right and didn't think it was plausible. But in the end, it became clear the problem wasn't with the data reduction—the problem was with the universe.”

  Or was it? All these grandiose conclusions—the runaway universe, expansion without end, and a world where Lambda or some other form of dark energy, not matter, dictates our ultimate fate—were hanging on observations of a handful of supernovas. In Perlmutter's set of forty-two supernovas, the average one was too faint by 0.3 magnitude, or about 25 percent. Considering past errors in cosmological measurements, that sounds like an awfully small effect from which to draw giant inferences. Hubble's first distance estimates, after all, were nearly ten times smaller than the modern values. Both the Supernova Cosmology Project and High-Z teams claim they are 99 percent confident that the brightness discrepancy is real, but they grow cagier when asked if the discrepancy proves we live in an accelerating universe. Their findings rely on a tremendous leap of faith in the knowability of the universe and in the reliability of their understanding of it. James Peebles at Princeton, who has watched lots of cosmological fads come and go over the years, initially said, “I would give better than fifty-fifty odds of Lambda having been detected. But I wouldn't give better than three-to-one.”

  For one thing, nobody really understands how Type 1a supernovas work. When astronomers look at distant supernovas, they are seeing them as they were billions of years ago when they lived and died; it has taken all this time for their light to reach telescopes on the earth. Did those ancient stars behave the same way as the much better-studied nearby ones? Might the stars themselves have been fainter in the past? Perlmutter's and Schmidt's teams have buried themselves in supernova research to make sure Mother Nature wasn't fooling them. But supernovas are incredibly powerful, intricate things. The closer scientists look, the more mysterious they seem.

  The detailed saga of a Type 1a supernova goes something like this. A modest, sunlike star in a double-star system grows old. It puffs up to become a swollen red giant star and then blows off its distended outer parts. The hot, extremely dense cinder left behind is the white dwarf. It is composed mostly of carbon, nitrogen, and oxygen, nuclear by-products from the fusion reactions that powered the star. Because its mass is so tightly packed, the dwarf generates an intense gravitational field that can strip gas from the other star. It continues to gobble greedily until it porks up to a mass of 1.4 times that of the sun. At that point, the dwarf's gravity abruptly overwhelms the electrical forces between electrons and protons, which previously kept it from collapsing further. In an instant, the star implodes and grows furiously hot. A wave of nuclear fire tea
rs through the star and unleashes so much energy that the star blows to bits, flying apart at more than five thousand miles per second. This pyrotechnic cloud of radioactive debris produces the brilliant light of the supernova.

  Type 1a supernovas all start as the same basic kind of star—a white dwarf of exactly 1.4 solar masses—so intuition tells you that the explosions should all be more or less the same. But intuition is a shaky basis for publishing in the Astrophysical Journal, and the current theoretical models of supernovas have more holes than a prairie dog town. First off, the details of what goes in that nuclear flame are far too complicated to simulate. Kirshner launches into a sardonic David Letterman-style monologue as he pokes fun at how little astrophysicists know about these details. “It happens in a flame whose thickness is millimeters, and the amount of burning depends on the topology of the flame. Give me a break. Can they calculate this? Nooooo!” he says, laughing. Furthermore, there is a woeful lack of information to feed into the theories. No one has ever observed the star that becomes a Type 1a supernova before it pops off, so all the ideas about them are inferences. It is like trying to figure out how to build an A-bomb from watching footage of a nuclear test.

  Even if the laws of physics haven't changed, the stars themselves have. Billions of years ago, the chemical makeup of the universe was different from what it is today. Galaxies had not been so thoroughly polluted by the heavy elements that spew from giant stars, novas, and supernovas. The stars within those distant galaxies were also different, presumably. So Perlmutter and Schmidt have studied supernovas in many astronomical settings—in youthful spiral galaxies, where new stars still are forming, and in mature elliptical galaxies, for instance. On the whole, the explosions all seem the same. But Adam Riess at Berkeley, working with Schmidt, spotted a possible sign of trouble. He found that distant explosions reach their peak noticeably faster than nearby ones. Such hastiness doesn't necessarily mean the distant ones were fainter, but it does suggest that astronomers cannot safely assume that Type 1a supernovas all live by the same rules.

  The other worrisome source of error is almost insanely simple: dust. Maybe the only reason the distant supernovas look fainter is that their light is absorbed by intervening, opaque material. Perlmutter and Schmidt have done everything they could to correct for this effect. Small dust particles scatter blue light more than red, so dust should redden the apparent colors of the supernovas. Despite careful studies, neither team sees any sign of such color changes. Of course, nature could be perverse and have filled the universe with a different kind of dust that scatters all colors equally. Such dust would be very hard to spot. If you had been wearing a pair of sunglasses all your life, you might never realize that you weren't seeing the world's true appearance.

  One way to address these concerns is to look at extremely distant supernovas. Up to a distance of about six billion light-years, Lambda makes supernovas fainter, because space has been expanding more quickly than expected. At greater distances, the effect changes direction. We're now seeing back to a period in cosmic history when Lambda has not had as much time to take effect and the universe is expanding more slowly than expected. In other words, the most remote supernovas should actually be brighter relative to their redshifts than they would be without Lambda. If dust is blocking the light, or supernovas were fainter long ago than they are today, the effect should keep getting worse as astronomers look over greater distances, further into the past.

  In October 1998, Perlmutter's team found one such distant supernova, nicknamed “Albinoni,” using the Keck II telescope, one of the few large enough to peer out to such distances. The name comes from Perlmutter's fondness for classical music and aversion to the usual impersonal number-and-letter designations that have come to dominate astronomy. Albinoni lies nearly ten billion light-years from the earth and appears a little on the bright side, just as Perlmutter hoped it would. At the January 1999 American Astronomical Society meeting, on the first anniversary of the momentous announcement, Perlmutter addressed a packed auditorium and sported a hopeful grin when he pointed to the dot on his graph indicating Albinoni. He saw it as one more sign that the big bang is speeding up. In April 2001, Riess analyzed an even more distant Type 1a captured by the Hubble Space Telescope. This star, too, appeared brighter rather than dimmer. Even doubting Thomases began to believe in dark energy.

  In fact, most astronomers embraced the new gospel almost immediately. “It's remarkable how little intelligent criticism there's been,” says Kirshner. He notes that the most intensely hostile comments have come from within the two supernova groups. For a while he told his fellow team members, “In your heart you know it's wrong.” Yet the outside reception could hardly have been warmer. The current generation of astronomers has grown up with competing cosmologies and with constant reminders that attaining cosmic enlightenment means recognizing there is more to the universe than meets the eye. Perlmutter and Schmidt were confused, but even more they were elated. The whole reason for embarking on such a grueling endeavor was to find something inexplicable.

  Even more striking was the reaction of the theoretical community, the deep thinkers of sci/religion who follow in Einstein's tradition of building the universe with the brain, an old envelope, and a good ballpoint pen. The supernova results brought them some incredibly good tidings. For years the observers had been telling the theorists that the density of the universe is much smaller than required by inflationary theory, which firmly predicts that the shape of space is flat and hence the density must lie exactly at the critical point between permanent expansion and eventual collapse. Surveys of the distribution and dynamics of clusters of galaxies put the true matter density much lower, at something more like 0.2 or 0.3 of that critical amount. Then along came Perlmutter and Schmidt, whose findings say there is exactly enough vacuum energy to place the total cosmic density where the inflationary cosmologists predicted (and hoped) it would be. “I thought those guys would take great pleasure in this, say, 'We have been right all along, this is the proof.' But it's not true,” Kirshner says with a bemused laugh. Guth listened to back-to-back presentations by Kirshner and Perlmutter at the 1999 American Astronomical Society meeting and reacted with a shrug. “It doesn't change things much for inflation,” he said nonchalantly.

  Einstein's followers adored the beauty of an inflationary universe whose density is very close to the critical value. They felt inflation had to be true, just as the great prophet felt light had to bend around the sun. “If you have a good theory, you pursue it until the data rule it out—and inflation was a good one,” says Turner. By the early 1990s, a number of theoretical cosmologists had decided that they needed Lambda back in their equations. It was the only thing that made their otherwise beautiful readings of cosmic scripture make sense. “Everyone was holding their noses, because of the checkered history of Lambda,” says Turner. “There were a few rats that jumped ship, but most of the rats stayed on board.” When Perlmutter and Schmidt spread the word that they had found signs of Lambda, Turner was pleased but not exactly astounded. For nearly ten years the theorists had known that their models worked best when cooked up with a dash of Lambda.

  But this was more than blind faith. Lambda had changed a great deal since 1917. When Einstein conceived of it, Lambda was purely a philosophical invention. Ever since Yakov Zeldovich explored the connections between cosmology and quantum physics, however, it had become an increasingly testable piece of the overall sci/religious description of the world. The virtual particles that constantly pop in and out of empty space, bizarre though they may seem, have measurable effects. Since the mid-1990s there's been another tangible and very persuasive argument in favor of Lambda. When Guth and Linde originally assumed that the universe had exactly the critical density, there were no observations to back them up. Now there are.

  Once again, the most decisive evidence in cosmology comes from studies of the cosmic microwave background, the echo of the big bang. Cosmologists have reconstructed the early history of
the universe with great precision and concluded that this background dates from the time when the universe was four hundred thousand years old, when matter cooled enough to form atoms and suddenly turned transparent to radiation. It is fairly easy to calculate the size distribution of the structures that could have formed in that time. The overall mix of large and small lumps in the early universe, revealed in 1992 by the COBE satellite, accurately matches the pattern predicted by inflation. Moreover, the largest structures serve as giant measuring sticks in the sky. If the geometry of the universe is flat, they should show up in a microwave map of the sky as markings about one degree wide, or about twice the width of the full moon. If space is curved, the markings should look distinctly smaller or larger, depending on the type of curvature.

  Studies from the two balloon-borne microwave telescopes—BOOMERANG and MAXIMA—show a pattern consistent with a flat, critical-density universe. That finding offers more support for inflation. “There are no noticeable discrepancies,” Guth says. More recently, the two balloon experiments have found secondary fluctuations in the microwave background that look like the effects of hot matter flowing into and sloshing out of dense regions in the early universe. These fluctuations imply a balance between matter and dark energy that exactly matches the supernova results.

  As a result of these developments, Lambda is not only acceptable, it is downright fashionable. Cosmologists no longer need to whisper that they are seeking the secrets of the Old One. Yet the quest toward cosmic truth is far from over. Einstein didn't know what Lambda was when he stuck it in his equations, and we still don't know. Cosmologists talk in vague terms about how Lambda might result from the vacuum energy predicted by quantum physics. But as physicists have noted for a quarter century now, quantum theory calls for a vacuum energy far, far beyond what would be contained within Lambda. So much energy would literally blow the universe to bits. That leaves Cosmologists in the awkward position of arguing that there is some mechanism, entirely mysterious, that cancels out all but one part in 10120 of the quantum vacuum energy, leaving just enough to make the current picture of the universe work properly. That is why Guth initially found the supernova results shocking—not because they indicate Lambda exists, but because Lambda is so small compared to the proto-Lambda that powered inflation. “We've got to learn more about what this dark energy is. There's nothing more fundamental than figuring out the energy that dominates the universe,” Perlmutter says.