A Cosmic Crisis? Dark doings in the universe Astronomers appear to have a heavenly crisis on their hands, and it concerns material they can't even detect. Professional star watchers thought for years that they understood the basic theory of how structure-galaxies and galaxy clusters-arose in the universe. Now, some are worried that they don't have it quite right. The recent observations that roused their concerns, however, also are providing data that are beginning to put a face on the mystery material that underlies the problem. That stuff is called dark matter. As researchers piece together a profile of this so-far elusive substance, which they believe makes up most of the universe's matter, they are starting to find out how it links the smallest structures in the universe to the largest. While physicists continue to search for dark matter particles using accelerators and underground detectors, astronomers have now joined the hunt. Researchers first proposed the existence of this ghostly material in the 1930s. That's when Fritz Zwicky noticed that galaxies in the Coma cluster were spinning so rapidly that all the visible material wasn't enough to keep them from flying apart. Some unseen matter, it seemed, had to be supplying the extra gravitational glue. Over the years, incentive to believe in this mystery material has only grown. In the late 1970s, for example, researchers measured the velocity of the outer parts of several galaxies. In galaxy after galaxy, they found that the outer regions rotated so fast that it was a wonder any galaxy was still intact. Once again, the laws of physics seemed to dictate that some unseen matter resides there and provides the missing gravity. Further evidence for dark matter comes from measurements on a more cosmic scale in the 1980s and 1990s. Using remote quasars as flashlights whose beams pass through primordial hydrogen clouds on their way to Earth, astronomers have measured the amount of deuterium-a heavy isotope of hydrogen-that formed when the universe was very young. This measurement is supremely important because from it, researchers can infer the cosmic abundance of baryons, which include the protons and neutrons that make up all atomic nuclei. That exercise has led astronomers to calculate that baryons account for less than 5 percent of all the matter in the universe. The rest must be some sort of exotic material that no telescope can see. Indeed, astronomers have come to think of luminous galaxies as mere bright flecks embedded in a halo of dark material. In the prevailing theory of dark matter, the mystery material has a one-dimensional personality. This type of dark matter, known as cold dark matter, would consist of sluggish particles that exert a gravitational tug but exhibit no other distinguishing feature. These particles would give off no light and would interact with each other only slightly, through the weak nuclear force-the same force that governs, for example, the radioactive decay of atoms. Because these putative particles move slowly, they would have clumped together earlier in cosmic history than did baryons. Therefore, dark matter would have provided the gravitational scaffolding necessary for the first galaxies to coalesce when the universe was less than a billion years old. In that respect, the cold-dark-matter model has proved remarkably successful at generating the kinds of large-scale structures seen in the universe today. When cosmologists apply the model to the finer scales of galaxies and smaller objects, however, the theory seems to run into trouble. Computer simulations of cold dark matter create universes that are far lumpier on these smaller scales than the real universe appears to be. The model predicts, for example, that the cores of galaxies are much denser than recent, high-resolution observations indicate. It also holds that dwarf galaxies, like the satellite galaxies orbiting the Milky Way, should be 100 to 1,000 times more numerous than astronomers have detected. There are other conflicts. According to the standard cold-dark-matter model, the smallest galaxies were the first to form, coalescing at a time when the expanding universe was younger and denser than it was when gravity later pulled together the more massive objects. It follows that dwarf galaxies should contain a higher density of matter than the others. Yet in reality, many dwarfs are no denser than other galaxies and much larger objects, such as galaxy clusters. Furthermore, several observations hint that the distribution of dark matter in galaxy clusters is spherical rather than football-shaped, as models of cold dark matter suggest. Some researchers, such as Christopher S. Kochanek of the Harvard-Smithsonian Center for Astrophysics in Cambridge, Mass., argue that many of the apparent points of conflict between theory and observation may vanish when cosmologists develop more sophisticated models for the complex effects that baryons have on galaxy formation. Unlike dark matter, baryons radiate light and exert pressure, and most computer simulations of cosmic evolution don't accurately incorporate these properties. Other researchers say that the apparent problems with the theory of cold dark matter are signs of a real crisis. "If we only had one problem to worry about, you might blame it on [modeling], but when you have five problems, it's not so easy to dismiss them," says Paul J. Steinhardt of Princeton University. No quick resolution Theorists have developed two main approaches to resolving the cold-dark-matter conundrums. Each of these alternatives invokes a different version of dark matter. Last month, astronomers working with NASA's Chandra X-ray Observatory reported new findings that could rule out one of these. The findings suggest, however, that the dark matter crisis may not be resolved any time soon. Astronomers are looking to the Chandra observations, along with a host of other ongoing studies, to reveal what dark matter is-and what it isn't. At stake, notes Steinhardt, isn't just a deeper understanding of cosmic structure. The identity of dark matter must fit with scientific understanding of the fundamental forces of nature: electromagnetism, gravity, and the strong and weak nuclear forces, he says. Supersymmetry, the leading theory to unify those forces, includes several elementary particles that make good candidates for dark matter particles. These particles would interact only through the weak force. That's a plus for the cold-dark-matter theory but may be problematic for the alternatives. In one of the alternative models, researchers including Craig J. Hogan and Julianne J. Dalcanton of the University of Washington in Seattle propose that dark matter particles are neither cold and sluggish nor hot and speedy. Rather, they are just warm enough to slightly resist the mutual gravitational attraction that brings them together. This resistance could have made the first clumps of matter that coalesced in the universe slightly puffier than they would be in the cold-dark-matter model, says Dalcanton. Since these clumps formed the seeds from which bigger structures arose, the puffiness could explain why dwarf galaxies aren't as dense as cold-dark-matter theory says they should be, she adds. Because of their higher temperature, particles of warm dark matter move faster than particles of cold dark matter. That motion might enable these particles to avoid congregating at the centers of galaxies. This would fit with the observed low density of galaxy cores. One possible strike against warm dark matter is described in a paper to appear in the Astrophysical Journal. Rennan Barkana of the Canadian Institute for Theoretical Physics in Toronto, Zoltan Haiman of Princeton University, and Jeremiah P. Ostriker, now at the University of Cambridge in England, note that the material's resistance to clumping might delay the early epoch when the very first quasars-and the supermassive black holes thought to power them-came into existence. In another version of the dark-matter theory, the mystery material, known as self-interacting dark matter, remains cold but is a lot more sociable. As proposed by Steinhardt and his Princeton colleague David N. Spergel, the particles interact strongly with each other, colliding and scattering like billiard balls. As with baryons, the collisions would occur more frequently in crowded quarters, such as the cores of galaxies, than in the sparse expanses of intergalactic space. In the simplest model, all dark matter particles would have the same likelihood of colliding, regardless of their speed. The jostling of self-interacting dark matter particles would tend to spread out the galactic cores, reducing the density there. Farther from these cores, in less compact regions, the particles would rarely meet and so behave like particles in the standard cold-dark-matter theory. Self-interacting dark matter could also explain the relative dearth of dwarf galaxies-or at least why so few are found buzzing around large galaxies-says Steinhardt. If there were interparticle collisions, the halo of dark matter surrounding a big galaxy would have a more pronounced tussle with the halos of nearby dwarf galaxies. The interactions would strip the dwarfs of their gas and stars more rapidly than in the standard cold-dark-matter theory. So, more of these dwarf galaxies would boil away or fall apart. Mapping dark matter Observations with the Chandra X-ray Observatory, reported last month in Washington, D.C., seem to have dealt a blow to the self-interacting model. To test the model, researchers used Chandra's sharp optics to measure the temperature and intensity of the hot, X-ray--emitting gas in a cluster called EMSS 1358+6245, which is some 4 billion light-years from Earth. Just as lights on a Christmas tree outline its dark branches, the X-ray-emitting gas provides a map of the dark matter in the cluster. With these data, John S. Arabadjis and Mark W. Bautz of the Massachusetts Institute of Technology, along with Gordon P. Garmire of Pennsylvania State University in State College, found that the density of the dark matter is greater the closer it is to the center of the cluster. Chandra could probe no closer than 130,000 light-years from the center, a distance much greater than the radius of an individual galaxy's core. Nevertheless, the findings still rule out the simplest model of self-interacting dark matter, Arabdjis' team says. "What we're seeing is the farther we go, the denser [the dark matter] gets," says Bautz. That's in contradiction to the self-interacting dark matter model, in which the particles bump into each other and keep the density from rising by puffing up the core. "So, our data completely support the standard picture [of cold dark matter]," says Bautz. Ostriker notes that having data from a single cluster isn't enough to knock down the self-interacting theory, but he says that further observations "can potentially provide a clue about what the dark matter is." Bautz agrees. "We're not saying that we now understand something about dark matter that we didn't before, but we're undoubtedly going to know more when all the Chandra data are in," he says. In some models of self-interacting dark matter, adds Ostriker, the force between the particles declines dramatically with speed. That's a crucial feature because the greater gravity in a galaxy cluster makes particles there move faster than they do in an individual galaxy. Consequently, self-interacting dark matter particles may have substantial collisions in a galaxy but act in a galaxy cluster just as inertly as do cold-dark-matter particles. The Chandra observations can't rule out that possibility, notes Bautz. Nor do they rule out warm dark matter. At the cores of galaxies, the faster-moving particles of this version of dark matter could offer some resistance to gravity, preventing the dark matter from congregating there. However, warm-dark-matter particles would be no match for the gravity of galaxy clusters. Warm dark matter would therefore behave no differently than cold dark matter in such a weighty environment. Prying into secrets The Chandra observations have extended the search for dark matter-once limited to particle accelerators and underground detectors-into the realm of astrophysical observations, says Steinhardt. With longer-term observations, Chandra will be able to peer even more closely into the centers of galaxy clusters and place new limits on models for dark matter, adds Bautz. Already, the Chandra observations are prying into dark matter's secrets. By placing limits on the strength of the interaction between dark matter particles, the results suggest that if the particles do collide, they do so relatively weakly. Several other astrophysical studies may also illuminate the dark matter mystery, says Steinhardt. For instance, astronomers can measure the density of small galaxies or the cores of larger galaxies by determining how well they act as gravitational lenses. Any dense object serves as a lens. It bends the light passing by it from a background body, such as another galaxy, into multiple images or arcs. The higher the density of dark matter, the greater the distortion. Since some dwarf galaxies may be essentially starless, and so all but invisible, the only way to detect them is through their distortion of the images of background objects. Gravitational lensing thus provides an accurate count of dwarf galaxies in a given patch of sky, a critical number for testing the predictions made by different dark matter models. Closer to home, increasingly detailed maps will provide an accurate estimate of the abundance of dwarf galaxies near the Milky Way, Steinhardt says. Their distribution provides another hint about the nature of dark matter. The cold-dark-matter theory predicts that dwarfs would be randomly distributed throughout the universe, but the self-interacting model suggests that relatively few should lie near big galaxies like the Milky Way. In contrast, the models indicate that warm dark matter would reside in sheets. More information about the nature of dark matter may come from the abundance of tidal tails, the streams of stars and gas that are gravitationally torn by the Milky Way from small galaxies, such as the nearby Large Magellanic Cloud. Self- interacting dark matter would hasten the stripping of these satellites, increasing the number of tails. It would also shrink the size of these satellite galaxies. "The exciting thing about this is that the realm of local astronomy is a new window on the nature of dark matter," says Steinhardt. "We're not talking about measuring distant galaxies but rather measuring satellites in our neighborhood and the neighborhood of the [nearby] Andromeda galaxy." With these studies, he says, the dark matter crisis may ultimately be resolved.