Gravitational wave discovery gives way to Milky Way dust

Planck spacecraft shows no evidence to support BICEP2 results

dust map of Milky Way

DUST MAP  In this map of the entire sky, Planck measured regions heavily contaminated by dust (red) and relatively clean ones (blue). The map looks above (left) and below (right) the plane of the Milky Way, centered on the galactic pole, with larger circles representing lower galactic latitudes. The black outline at right shows the sky region in which BICEP2 looked for gravitational waves.

Planck Collaboration, adapted by E. Otwell

New data from the European Space Agency’s Planck satellite spell more trouble for the claimed discovery of ripples in the fabric of space created moments after the Big Bang. The Planck data strongly suggest that dust in the Milky Way galaxy might account for the entire signal interpreted as gravitational waves by researchers using BICEP2, the Antarctic telescope responsible for the initial discovery, which was announced in March.

If the BICEP2 observations hold up, they would be the first direct peek at the long-hypothesized epoch of inflation, a period of explosive cosmic expansion that followed the birth of the universe (SN: 4/5/14).

Over the months since the announcement, however, doubts have surfaced. Many researchers wondered if the BICEP2 team had properly accounted for the amount of dust in the Milky Way, which might interfere with the observations (SN: 6/28/14).

Planck’s results, reported Sept. 21 at arXiv.org, strongly suggest that BICEP2 didn’t see gravitational waves, only dust in our galaxy.

“There is more dust in the BICEP2 signal than they accounted for,” says Jan Tauber, a European Space Agency astronomer and Planck team member. But, he notes, there’s enough uncertainty that gravitational waves might still lurk in the data.

Both BICEP2 and Planck observed radiation from space known as the cosmic microwave background — the faint glow of the first light released into space as atoms formed out of the primordial fog, 380,000 years after the Big Bang. BICEP2 looked for gravitational waves by hunting for twirling patterns imprinted on the alignment, or polarization, of this microwave light. The telescope stared at one patch of sky for nearly three years.

Unfortunately, interstellar dust — sootlike grains of carbon and silicon — can mimic the polarization pattern of gravitational waves. BICEP2 researchers took this into account by relying on six estimates of Milky Way dust. They also picked a part of the sky where dust should be relatively sparse.

BICEP2 is more sensitive than Planck but it measured light at only one frequency, 150 gigahertz, where dust is hard to detect. Planck mapped the polarization of the entire sky at seven frequencies, many of which are more sensitive to dust. Those higher frequencies provided the first direct measurement of dust polarization over the entire sky.

“This is a significant change,” says Lloyd Knox, a cosmologist at the University of California, Davis, who also works with the Planck team. BICEP2 had to rely on calculations to guess the interference from dust, he says. “Now there’s a much better estimate of the contamination solidly grounded in data.”

Jamie Bock, a Caltech cosmologist and one of the leaders of the BICEP2 team, admits that the initial analysis probably overestimated the strength of gravitational waves. “The dust level is significant,” he says. But it’s too early to know whether dust makes up the entire signal. “The analysis is not a one-to-one comparison with the signal reported by BICEP2,” he says.

That’s because the instruments on BICEP2 and Planck are very different, complicating a direct comparison. Also, Planck’s interpretation of the BICEP2 observation relies on extrapolating from observations of the sky at 353 gigahertz down to BICEP2’s frequency of 150 gigahertz. The extrapolation is guided by observations at intermediate frequencies of the entire sky. But there’s no guarantee that dust seen by BICEP2 behaves the same as dust from other parts of the sky.

In July, the teams announced that they would share data to help resolve the controversy. The teams plan to publish that analysis in late November.

Meanwhile, many cosmologists, including members of both BICEP2 and Planck, emphasize that the current maps are not the final word. “The results are not definitive,” says Knox.

Scott Dodelson, a cosmologist at the Fermi National Accelerator Laboratory in Batavia, Ill., agrees. While the Planck results imply dust is the culprit, he says, “there’s lots of room to go one way or the other.”

Planck discovered that no part of the sky is free from dust, but some patches are cleaner than the one chosen by BICEP2. “This will affect strategies for the future,” says Knox, for the many other experiments hunting for gravitational waves.

“These data are invaluable,” says William Jones, a Princeton cosmologist who is in charge of SPIDER, a balloonborne experiment designed to hunt for gravitational waves with observations at two frequencies. The balloon will launch over Antarctica in December.

The Planck maps, he says, will help the SPIDER team plan observations and sample many patches of sky with minimal dust. If the data from different parts of the sky agree with one another, Jones says, then they probably have a common origin in the cosmic microwave background. Then the team can be confident that SPIDER is seeing gravitational echoes from the birth of the universe.

Editor’s Note: This article was updated September 26, 2014 to correct the time of the planned launch and number of frequences of the SPIDER experiment.

Christopher Crockett is an Associate News Editor. He was formerly the astronomy writer from 2014 to 2017, and he has a Ph.D. in astronomy from the University of California, Los Angeles.

More Stories from Science News on Cosmology

From the Nature Index

Paid Content