In the late 1990s, two teams of astronomers stunned the scientific community with the finding that the universe is accelerating in its expansion, somehow overpowering the constant pull of gravity that should be slowing it down. The culprit pressing the cosmic accelerator goes by the name "dark energy," which is an appropriately enigmatic moniker for something that remains so poorly understood.

"We have an amazingly simple picture of the universe," says Princeton University astrophysicist Michael Strauss. "Of course, we don't understand that picture—we don't know what dark energy is, and we don't know what dark matter is." Dark matter, a mysterious entity of longer standing, is some invisible but common substance that reveals itself only through its gravitational pull.

But dark energy—whatever it is—is there, according to a number of measurements taken in the years since its influence was first detected. Now a pair of researchers at the University of Provence in France has added to the body of evidence by confirming dark energy's presence through an independent test that verifies the impact of cosmic parameters on the appearance of pairs of distant galaxies. The research appears in the November 25 issue of Nature. (Scientific American is part of Nature Publishing Group.)

Determining where a distant object lies in space takes a bit of work, because the universe lacks clear distance markers. Astronomers and cosmologists can infer the three-dimensional position of a star or galaxy by measuring its redshift, which reveals not the actual distance to the object but how much its emitted light has been stretched by its recession from us within an expanding universe. Then, with a few assumptions about the curvature and contents of the universe, they can reconstruct the positions of those objects from redshifts. (Space can have positive curvature, like the surface of a sphere, or negative curvature, like the surface of a saddle. Only in a flat universe does space obey all the standard geometry-class rules—the angles of a triangle add up to exactly 180 degrees and parallel lines never meet.)

Decades ago, researchers realized that if they could observe some spherical distribution of objects in the distant universe, they could use any apparent distortion in that sphere to determine the universe's geometry and contents. After all, only with the correct parameters that convert redshift to position would the reconstructed distribution be spherical. The test was proposed in 1979 by Charles Alcock, now at the Harvard–Smithsonian Center for Astrophysics, and the late Bohdan Paczynski, who was a professor at Princeton until his death in 2007. The researchers "pointed out that if you're looking at the distribution of galaxies, and you're looking at the wrong cosmology, you're going to end up with a distorted mess," Strauss says. "Things that would otherwise be round look distorted."

University of Provence cosmologist Christian Marinoni and graduate student Adeline Buzzi, the study's authors, took a new approach to the Alcock–Paczynski test, concentrating on the individual alignment of hundreds of galactic binaries. The orientation of those gravitationally bound pairs of galaxies should be completely random, as viewed from our vantage point within the solar system. "Those galaxies have no idea that you're there watching them, so it's just some random variable," Strauss says. Imagine two points on opposite sides of a sphere—rolling the sphere around reveals all the different orientations a pair of galaxies might take for any particular observer, from a head-on arrangement to a vertical stacking to any flavor of tilt.

But the geometry and expansion of the universe can distort the apparent orientations; without the proper corrections for the universe's makeup and shape, the orientation of galactic binaries will look warped. "The apparent orientation is biased because we measure orientation not with a compass or with a ruler but with redshift," Marinoni says. And redshift depends on just how the universe is expanding.

By tweaking the universe's geometry and the nature of its dark energy, the researchers corrected the picture until the galactic couples were indeed pointed in all directions, as would be expected. With those tweaks, Marinoni and Buzzi confirmed two tenets of the current cosmological model: that the universe is a flat space and that it is dominated by a dark energy, which makes up roughly two-thirds of the universe, that looks a lot like Albert Einstein's famed cosmological constant. (The rest comes primarily from dark matter, with ordinary matter—atoms and molecules—contributing just 4 percent or so to the total makeup of the universe.) "You have a distorted image of these couples, but when you put in the good, flat curvature of the universe, and the good amount of dark energy, then immediately you recover the isotropic [symmetrical] arrangement of these couples," Marinoni says.

The new twist on the Alcock–Paczynski test was not simply a creative leap—Marinoni and Buzzi also had to correct for some pesky redshift effects that come from the galaxies' own velocities, independent of the universe's expansion. Marinoni likens the process to clocking a speeding car on an expanding street; cosmologists want to know not how fast the car is moving on its own but how fast the spreading street is carrying it along. Without correcting for the galaxies' own motions, which are known as peculiar velocities, the binary pairs tend to appear more elongated along an observer's line of sight.

So the researchers measured the orientation of 721 nearby, or low-redshift, galactic binaries in archival data from the Sloan Digital Sky Survey, calibrating them to account for the contribution of the galaxies' own velocities. Armed with the assumption that nearby galaxy pairs move in the same way as those in the distant universe, the researchers applied their calibration to 509 faraway, or high-redshift, galactic binaries from the DEEP2 redshift survey to isolate the true orientation of the distant pairs.

But that assumption leaves a little daylight in the new case for dark energy. If the distant galaxies have different peculiar velocities than the nearby galaxies, the researchers' results would be skewed. Still, the test is a new look at the curious phenomenon of dark energy, and its findings agree well with a mounting body of evidence from different perspectives. "Thus far, the picture has been pretty rosy, in that all the tests that have been done seem to fit together," Strauss says. "But the more tests you have, the better off you are."