The controversy continues to develop over the interpretation of the results from BICEP2, the experiment that detected “B-mode” polarization in the sky, and was hailed as potential evidence of gravitational waves from the early universe, presumably generated during cosmic inflation. [Here’s some background info about the measurement].
Two papers this week (here and here) gave more detailed voice to the opinion that the BICEP2 team may have systematically underestimated the possible impact of polarized dust on their measurement. These papers raise (but cannot settle) the question as to whether the B-mode polarization seen by BICEP2 might be entirely due to this dust — dust which is found throughout our galaxy, but is rather tenuous in the direction of the sky in which BICEP2 was looking.
I’m not going to drag my readers into the mud of the current discussion, both because it’s very technical and because it’s rather vague and highly speculative. Even the authors of the two papers admit they leave the situation completely unsettled. But to summarize, the main purpose and effect of these papers seems to be this:
- to point out that the level to which dust tends to be polarized, as measured by the Planck satellite team in their very recent paper, is larger than most people expected: twice as large, or more, than assumed by BICEP2 in some of their arguments;
- to remind us that the effect of the dust on the BICEP2 polarization signal increases like the square of the dust polarization — i.e., a doubling of the dust polarization means a quadrupling of the contribution of the dust to BICEP2’s measurement;
- to try to convince us that the uncertainties about dust are currently too large for anyone to draw conclusions as to the meaning of BICEP2’s measurement — as to whether it is due to gravitational waves, to polarized dust, or to a combination of the two [along with small effects from synchrotron radiation and from gravitational lensing of E-mode polarization of the cosmic microwave background];
- to note that new measurements relevant for determining the dust effect are on their way on the 6 to 12 month timescale, so these issues may soon be resolved — although it may not be so easy, because precise determinations of the dust polarization, and proper combination of measurements by different experiments, will be needed for an unambiguous measurement of any gravitational wave contribution to BICEP2’s B-mode polarization.
Regarding the last point, the most important measurements in question will come
- from the BICEP2 team itself, whose next-generation experiment KECK, already running, is basically five BICEP2’s in tandem, and will cover two different microwave frequencies, and
- from the Planck satellite at a range of frequencies.
Both papers make some basic points that undermine confidence in BICEP2’s case against dust (e.g. by arguing that BICEP2, by overestimating the intensity of dust emission as measured by Planck, was led to strongly underestimate the average amount of dust polarization; that BICEP2 disregarded lensing effects in one of its arguments against dust; that BICEP2’s evidence that their data was uncorrelated with dust measurements carries no weight). But when the papers try to make more quantitative statements about just how big the dust effect could be… well. As far as I can tell, the methods used involve arbitrary judgment calls, and in the second paper these aren’t even clearly explained, and lead to key claims that can’t be verified by the reader. That said, I’m not suggesting that anyone could do any better right now, given how little information we have about polarization in the BICEP2 region. But I would suggest that experts read the words in these papers very carefully before attempting to evaluate the plausibility and reliability of the most important plots shown.
Since the whole discussion is currently so vague, I probably won’t say more about this until someone says something more convincing, or until BICEP2 publicly responds, if they choose to do so. The bottom line, for now, is clear enough: BICEP2 hasn’t made a case that overwhelmingly convinces the world’s experts. The community will therefore remain collectively undecided until a precise determination is made of the amount of polarized dust in BICEP2’s region of the sky. That lies six months to a year in the future. Until then, our knowledge will remain hazy.