Gravitational Waves and the Value of Errors
November 11, 2014
Imagine a fat guy doing a cannonball into a calm swimming pool. We can predict that a minute later the surface of the pool will be mottled with waves: ridges and valleys, peaks and pits. The Standard Model of cosmology makes a similar prediction, that the rapid inflation of space after the Big Bang left it mottled with gravitational waves where space itself expands and contracts, much like the movement of water molecules on the surface of the pool. We could sprinkle pepper on the pool to easily observe and measure the waves. Nature has already done this for us in space: the light from cosmic background radiation gets polarized as it passes through the peaks and pits of gravitational waves, allowing us to observe the waves indirectly.
Just as we can model the waves in the pool to work backwards and learn about the triggering splash, cosmologists hope that similar modeling of gravitational waves can provide new evidence of, and insight into, the earliest moments of the Universe. The applications of such knowledge are vast, and key to some of our most fundamental questions. Proof of these waves has been called the "Holy Grail of inflation."
This is why the science community was so excited when, in March of 2014, scientists operating the BICEP2 microwave telescope at the South Pole announced the discovery of polarized light matching the predicted patterns of gravitational waves after studying a specific patch of the sky for three years. They made their announcement only after having established that the chances a spurious background could have accounted for their data was less than one in a million. "Everything about the signal," wrote physicist Lawrence Krauss in Scientific American, "empirical confirmation or refutation of BICEP2 should be possible within a year or so."
The leading suspect for triggering such a false alarm is simple interstellar dust here in our galaxy. Vast clouds of carbon and silicon dust can also polarize microwave radiation passing through; and, inconveniently, these clouds come in just the right sort of sizes and shapes as as the peaks and pits of the predicted gravitational waves. Although the BICEP2 team has always been well aware of this, they had sufficient confidence in their data while admitting being unable to rule out the party-pooping dust. This is not bad science, it is the best science.
It was in September 2014—smack in the middle of Krauss's "year or so"—when the first strong evidence was reported that the BICEP2 observation was of light polarized by dust, not by gravitational waves. It came from the European Space Agency's Planck satellite, whose telescope is weaker than BICEP2 but is tuned to a wider range of microwave frequencies. By observing only specific frequencies, some of which are better able to detect the dust, Planck found that the polarizing dust is much more common than previously thought. According to the Planck team, the concentration of dust is more than sufficient to fully account for the BICEP2 observations, no gravitational waves needed.
But, continuing the "best tradition of science," the teams are now sharing their data directly to see whether a gravitational wave signal may still lurk within the polarization from interstellar dust. Neither the BICEP2 nor the Planck announcements brought a final verdict. Instead, they are building a knowledge base upon which future researchers may improve.
That very process is already underway. Planck produced a dust map of the entire sky, allowing future experiments to perform detailed studies in small patches of sky that are relatively free of interference from dust. One such team, SPIDER, is currently preparing to launch a balloon-borne telescope from Antarctica next month. It will use the lessons learned from both BICEP2 and Planck to make these detailed studies, on two frequencies, observing clear patches of sky, looking for light polarized by gravitational waves.
The scientific method often depends on the scrutiny of results once thought to be conclusive. If BICEP2 was right, we have learned something; but if it was wrong, we have learned two things. What's certain is that eventually, the BICEP2, Planck, and even the SPIDER data will be repeatedly re-examined and the lessons learned from them will improve our knowledge time and time again. Are there gravitational waves out there that can prove inflation took place? We don't know yet, but the scientific method will continue to nudge us ever closer.
@Skeptoid Media, a 501(c)(3) nonprofit