The Quest for Cosmic Inflation

Evidence of Inflation

Based on Einstein’s general theory of relativity, inflation would generate gravitational waves that squeeze space as they traverse, and this squeezing leaves a unique imprint on the CMB, the primordial afterglow left behind by the Big Bang. This faint afterglow exists even today as shown by Penzias and Wilson in the 1960’s and later more convincingly by the satellite data.

Indeed, researchers from the Background Imaging of Cosmic Extragalactic Polarization 2 (BICEP2) collaboration announced the first direct evidence of the cosmic inflation theory by finding a pattern called primordial B-mode polarization in the CMB. Because the CMB is a form of light, it exhibits all the properties of light, including polarization. Just like sunlight is scattered by the atmosphere and becomes polarized, so was the CMB scattered by atoms and electrons comprising the primordial universe and became polarized as well. B-mode is a special type of polarization, which represents a twisting or curling pattern in the polarized orientations of the ancient light. This pattern can only be created by gravitational waves produced by inflation 14.

The BICEP2 telescope is set up at the South Pole to take advantage of the cold, dry, stable air, which allows for precise detection of the faint cosmic light. The researchers were surprised to detect a B-mode polarization signal considerably stronger than what many cosmologists expected, consistent with a “5-sigma” statistical significance necessary for a true discovery (see Figure 12). The team analyzed their data for more than three years to eliminate possible errors. They also ruled out the contribution of dust in our galaxy that could presumably produce the observed pattern.

Figure 12
Figure 12. BICEP2 uses about 250 thumbnail size polarization detectors to look for a difference in the CMB light from a small patch of sky coming through its telescope in two perpendicular orientations. The instrument collected data between January 2010 and December 2012 at the Amundsen–Scott South Pole Station, where the cold, dry air offers especially stable viewing conditions.

The BICEP2 researchers have reported a surprisingly large number for the parameter r:

r=\frac{\mbox{gravitational wave fluctuations in the CMB}}{\mbox{fluctuations caused by perturbations in the density of matter}}

r was estimated to be less than 0.11 using the all-sky CMB maps from WMAP and the Planck satellites. BICEP2 measured this value to be around 0.20—a high value of r indicates that inflation began even earlier than what some models predicted: 10^{-36} seconds after the Big Bang.

The timing of inflation is important because it points to the energy scale of the universe when inflation was in progress. BICEP2’s value of r suggests that this was the same energy scale at which all the forces of nature except gravity (the electromagnetic, strong and weak forces) might have been unified into a single force—an idea called Grand Unified Theory (GUT). The finding strengthens the idea of grand unification and rules out a number of inflation models that do not feature such an energy scale.

It seems that finally, scientists can claim to understand what went on at the beginning of everything. But could they indeed? Things looked too good to be true.

Correctness of scientific work is validated by a peer review process. Often great discoveries are challenged with questions and doubted at the onset, which is the essence of the peer review process. Often, experimental results reported by a team of researchers cannot be reproduced and verified by other teams, thus the original results become untrustworthy and fade away into oblivion.

BICEP2’s results were quickly challenged by Planck scientists. Nobody in this case doubted that BICEP2’s telescope had detected B-mode polarization in the sky above South Pole, as the scientists appeared to have successfully eliminated systematic instrumental errors15. But whether BICEP2 indeed captured the evidence of this polarization prompted by primordial gravity waves became a matter of great contention.

In order to find out how significant the gravitational wave B-mode signals were, scientists needed to eliminate (a) the effect of intergalactic dust (dust B-modes) and (b) the gravitational lensing effect of massive galaxies (lensing B-modes) from their results. The BICEP group was able to eradicate the lensing B-modes but could not fully eliminate the dust B-modes because a reliable map of the dust’s contribution toward the total B-modes did not exist until the Planck group published their results much later. The Planck apparatus could survey the entire sky in multiple frequency regimes—specifically around 353 MHz, where the contamination of dust is the strongest (dust has distinct contributions at different frequencies). The BICEP2 scientists believed their B-mode signal came from a small patch of the sky where the effect of dust was not very pronounced. But they were wrong.

The BICEP2 scientists made significant improvements to their hardware since their ill-fated announcement. They added the Keck Array, a powerful suite of telescopes, and continue to measure the Cosmic Microwave Background polarization at high precision in search of the B-mode signature of inflation. But a joint analysis of data from ESA’s Planck satellite and the ground-based BICEP2 and Keck Array experiments found no conclusive evidence of primordial gravitational waves. Planck’s data showed the contribution of dust B-mode in the total B-mode signal is larger than what the BICEP2 team originally expected. When the well understood lensing B-mode and the newly uncovered dust B-mode components were subtracted from the data, there was hardly any signal left of gravitational waves etched into the CMB—definitely not enough to claim a discovery.

On February 3, 2015 BICEP2 scientists eventually withdrew their claim that they found the first evidence for the primordial “B-mode” polarization of the CMB.

It is important to note, however, that this null result does not disprove the existence of primordial gravitational waves. It just says that the original BICEP2 claim was flawed although there still is a possibility that the gravitational wave signal is hidden in the BICEP2 data, but drowned in noise of galactic dust. The new result placed an upper limit on the possible strength of B-mode polarization caused by gravitational waves.

In spite of the setback, continuing to attempt precision measurements of the CMB polarization is important as it is the oldest light in the universe and an extremely powerful tool that can be used to probe the nature of the universe at very early times. A number of satellite and ground-based experiments—such as LiteBIRD, COrE, Atacama Cosmology Telescope and the recently launched SPIDER telescope—are now being planned or are operational. Their goal is to measure the CMB polarization and settle (a) the debate surrounding primordial gravitational waves, and (b) the question whether cosmic inflation is a viable theory of the universe.

So the search for primordial gravitational waves (emitted during the cosmic inflation era around 10-38 second after the Big Bang) continues. It is an exciting time for astrophysicists. If we obtain r equal to 0 from these new measurements, then there were no primordial gravitational waves to begin with, and it will be difficult to make a case for cosmic inflation. Theoreticians will have to go back to their drawing boards to figure out what really happened at the beginning of the universe. On the other hand, large values of r could support the idea of grand unification of forces, when all the forces of nature except gravity (the electromagnetic, strong and weak forces) may have been unified into a single force.


14 Much like light waves, gravitational waves have a “handedness”, and can have left- and right-handed polarizations. The swirly B-mode pattern is a unique signature of gravitational waves because of their handedness.

15 Interestingly, BICEP2 did, in fact, take some of this foreground polarization into account by reverse-engineering data taken from a PDF slide of Planck results since Planck’s raw data had not been released at the time. The process is sometimes called “data scraping”. Admittedly, it is not an ideal way of data analysis, but often works moderately well. In fact, Planck foreground is one of several foreground effects that BICEP2 did account for but it is not clear if the dust noise is a large or rather a minor error.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s