Science 1663

• Los Alamos Science and Technology
  Magazine
  JANUARY 2007
• About 1663     |     Past issues

Lighting Up the Dark Universe

Los Alamos, a world leader in computing since the beginning of the digital revolution, is mounting a major initiative to help unravel the mysteries of dark matter and dark energy.

Shown here are two nearby galaxy clusters that collided about 150 million years ago. The gases from each cluster (red regions imaged with x-rays) were slowed down by collisions and left behind, whereas the massive dark matter components (blue regions imaged through gravitational lensing) are collisionless and therefore passed through each other unimpeded. This separation of dark and ordinary matter is being touted as the first direct evidence of dark matter. Photo credits

For most of the 20th century, everything that cosmologists observed in the heavens confirmed the laws of physics we know on Earth.

But in the background, a crisis has been building. Starting with early (1933) observations of galaxy clusters, evidence has accumulated to suggest that the matter in planets, stars, and interstellar gas—ordinary matter made of neutrons, protons, and electrons—is but a small fraction of the matter in the universe. Most of it appears to be cold and dark, to have no electric charge (making it unable to emit or absorb light), and to collide so infrequently with other matter that it never heats up or cools down.

This "dark" matter is invisible to us, but Newton's laws tell us that it must exist to provide the gravitational force that keeps the fastest stars confined within our own galaxy and the fastest galaxies bound into giant clusters.

Apparently, all the luminous matter we observe is embedded in massive amounts of unknown dark matter in the form of extended halos. That was hard enough to swallow. But things got even stranger.

In 1998, the search for distant supernovae (exploding stars) revealed that the overall expansion of the universe, which began about 14 billion years ago, is not slowing down as it should under the braking power of gravity.

Instead, the expansion appears to be accelerating under the influence of a mysterious force, dubbed dark energy. Over the past eight years, the evidence for dark energy has finally turned the world of physics on its ear. Computer simulations of the dark matter universe, when compared with the latest maps of luminous matter and the latest maps of the cosmic microwave background (the radiation left over from the early universe), indicate that the universe contains an astonishing 74 percent dark energy, 22 percent dark matter, and only 4 percent ordinary matter.

"This is a very stimulating time for physics. Fully 96 percent of the universe seems to be composed of stuff we've never seen directly on Earth!" says Emil Mottola, Los Alamos theorist, who has his own model of dark energy.

According to the Dark Energy Task Force, a joint committee sponsored by the U.S. Department of Energy, the National Science Foundation, and the National Aeronautics and Space Administration, "[N]othing short of a revolution in our understanding of fundamental physics will be required to achieve a full understanding of cosmic acceleration."

A Sloan Digital Sky Survey map of the luminous universe out to a distance of 5 billion light years. We (and the present time) are located at the center between the two pie-shaped regions. Note how galaxies condense into lace-like filaments and walls as time progresses from the outer edges of the map to the center. Photo courtesy of the Sloan Digital Sky Survey (sdss.org)

The three agencies have plans for a multi-billion-dollar exploration of the deep universe. By charting its expansion history back to the time when the universe was less than half its present size, the astro-explorers hope to track the ebb and flow of cosmic acceleration and how it has altered the rate at which galaxies merge into clusters, superclusters, and huge lace-like structures of filaments and walls (see the figure at left). If they can discover the true nature of dark energy, they will find out whether Einstein's theory of general relativity, the description of the expanding universe that has held for 75 years, needs to be changed in some fundamental way.

Success will depend on coordinating theory, computation, and the many types of observation planned for the future. Lab Director Mike Anastasio has named the problem a Grand Challenge for Los Alamos, and research scientists at Los Alamos are gearing up for the task by developing close collaborations with large astronomical surveys, including the Sloan Digital Sky Survey at the Sloan telescope in southern New Mexico. The Sloan survey is the largest sky survey ever undertaken.

Unbounded Acceleration?

The surveys hope to test various theoretical notions of dark energy. The simplest, called the cosmological constant, postulates dark energy as a constant energy density tied to empty space. Its presence is felt as a negative pressure, or repulsive force, opposing the force of gravity and putting a damper on the tendency of matter to clump.

Mike Warren, winner of four Gordon Bell prizes for his high-speed simulation code HOT, is providing catalogs of dark matter halos to many research groups across the country. Photo by John Flower

According to the theory, when the universe is young and dense, the dark energy has a very small effect, but as the universe expands, the total amount of dark energy increases to fill the volume, causing the expansion to accelerate faster and faster. In about 100 billion years, this strange energy density causes most of the luminous matter in the star-filled heavens, except for a few galaxies near our own Milky Way, to disappear from our view.

Albert Einstein's theory of general relativity permits empty space to exhibit this bizarre property and, because in 1917 scientists thought the Milky Way was the whole universe, Einstein invoked the cosmological constant to convert his prediction of an expanding universe into a description of a stationary one.

Later, after Edwin Hubble discovered that the universe was indeed expanding, Einstein openly expressed his disdain for the cosmological constant, calling it an ugly thing that should never be realized in nature.

Warren's billion-particle simulation showing the typical distribution of dark matter in the universe today. Dark matter density increases from blue to light blue, to yellow, to orange. Ordinary matter condenses into stars and galaxies in the higher-density regions. Photo by John Flower

"Today the cosmological constant is back and being used in computer simulations," says Salman Habib, a theorist at Los Alamos. "But it has no natural explanation at the quantum level and brings home the fact that we don't have a consistent quantum theory for gravity." Quantum theories have successfully described three of nature's basic forces (electromagnetic, strong nuclear, and weak nuclear) in terms of a duality between particles and waves. That same approach applied to gravity suggests that an energy density such as the cosmological constant could arise from energy fluctuations in empty space, that is, from particles that pop briefly in and out of existence. However, this approach also predicts a much larger value for the energy density than has been observed—larger by a factor of 3 followed by 121 zeros!

The contradiction has led theorists to posit new ideas of what dark energy might be, including quintessence, a dynamical fluid filling all space, and more-radical models that modify Einstein's theory of general relativity. "To begin to distinguish the various models," says Habib, "we need to reduce the overall uncertainty in theory and in observations from 10 percent to about 1 percent. This is going to be extremely difficult."

The density of dark energy is miniscule compared with the density of matter in regions like the solar system and our own galaxy. Only by observing huge volumes of space extending hundreds of millions of light-years across the sky (a light-year is 10 trillion kilometers) and billions of years back in time can astronomers hope to trace the imprint of dark energy on the dynamics of the universe.

Mapping the Universe

As part of its dark energy effort, Los Alamos has become a member of the Sloan Digital Sky Survey (http://www.sdss.org/).

Dedicated to mapping the universe, the Sloan survey has imaged more than 200 million celestial objects, and its researchers have seen back in time to when the universe was about 5 billion years old and only two-thirds of its present size. Situated at Apache Point in the Sacramento Mountains of southern New Mexico, where crystal clear nights make for ideal viewing of the distant universe, the Sloan 2.5-meter-diameter digital telescope records continuously through the night, imaging a narrow strip of sky the width of the moon as the Earth turns on its axis. The strips are then laid side by side to give a contiguous view of the quarter of the sky visible to the telescope.

The Sloan telescope with John Peoples, former survey director. Photo courtesy of the Sloan Digital Sky Survey (sdss.org)

"The Sloan survey was the coming of 'big science' to cosmology," comments Habib. "It's no longer a single astronomer observing the night sky, but large teams of people gathering the data and ensuring its precision. The follow-up requires state-of-the-art computing infrastructures to handle very large data sets and large-scale simulations to connect observations to theory. Los Alamos and other national laboratories provide an ideal place for this type of analysis."

Powerful computing capabilities, developed to simulate the performance of nuclear weapons in the U.S. stockpile, can be applied to the problem of simulating the cosmos. "We are also developing a unique statistical approach to minimize uncertainties in the predictions drawn from computer simulations," says Los Alamos scientist Katrin Heitmann. "It allows us to get more-accurate results with many fewer simulations and to interpolate to new models of what the universe might look like."

Los Alamos will build on an impressive track record in computational cosmology. Back in the 1990s, when theorists and observers were in strong disagreement about the dominance of dark matter, Los Alamos scientist Mike Warren used the computer to follow the expansion of the universe in one of the largest cosmology simulations ever done at that time, one based on 17 million particles of cold dark matter.

Warren's prediction about the average velocity of one galaxy relative to another was easy to check. At first, Warren's result was much higher than that reported by observers, but careful comparison showed that the observers had made an error, inadvertently removing the high-velocity galaxies of a very large nearby cluster from their published analysis. When those galaxies were included, observations agreed with simulations. Warren wears a sphinx-like smile as he quietly states, "We began to realize that the computer was becoming as important as the telescope in shaping our understanding of the universe."

Ten years later, Warren's original computer code, run on the Lab's giant supercomputers, can simulate the evolution of a billion particles in a volume 3 billion light-years on a side. It is being used by several groups to interpret the Sloan data, including the data showing huge matter-density waves with a wavelength of about 300 million light-years.

Warren's code is now one of many that predict the dark matter distribution using variable amounts of dark energy, dark matter, and normal matter and different rates of cosmic expansion. The dark matter distributions are then lit up by populating them with galaxies, and the resulting distributions of galaxies and galaxy velocities are compared with data from the surveys. The problem is that the codes don't all agree, and the observations may be plagued by systematic errors. If the fate of the universe is to be known, then astronomers, physicists, and computer wizards have their work cut out for them.

In the next 10 to 20 years, astronomers hope that complementary observations will circumvent the problem of systematic errors and lead to 99 percent overall accuracy. In the meantime Habib and colleagues hope to establish an international computational cosmology center at Los Alamos, where astronomers can analyze all the measurements and codes in a consistent way to find out whether or not the universe is undergoing unbounded acceleration.

Evidence for a dark matter halo between us and a quasar. An invisible dark matter halo acts as a gravitational lens, bending the light from the quasar and producing five quasar images (circled). Photo courtesy of NASA, ESA, K. Sharon (Tel Aviv University) and E. Ofek (Caltech).

Strong gravitational lensing. A dark matter halo (center) between a distant galaxy and the telescope bends the galaxy's light to create two distorted images "above" and "below" the galaxy's actual location. Photo courtesy of NASA, ESA, K. Sharon (Tel Aviv University) and E. Ofek (Caltech).

Others, like Mottola, are using pencil and paper to search for a new idea that will fill the void between Einstein's theory of gravity and the quantum theories of nuclear and particle physics. Mottola explains, "Dark energy may be telling us that quantum fluctuations have gravitational effects at the very largest scales. This would be our first real hint from experiment at what a quantum theory of gravity should be like, and it may lead to entirely new directions in physics. For example, quantum fluctuations could change present ideas about black holes, with a dark energy repulsive core replacing the singularities of Einstein's theory. New ideas are definitely needed."

These are heady times in physics, with the biggest questions of all—what is the universe made of and how did it get here—being asked once again.

Credits for photo one: X-ray: NASA/CXC/CfA/M.Markevitch et al.;
Optical: NASA/STScI; Magellan/U.Arizona/D.Clowe et al.;
Lensing Map: NASA/STScI; ESO WFI; Magellan/U.Arizona/D.Clowe et al.