In the Dark

I’m not meant to be blogging these days but I thought I’d break radio silence to draw attention to a new paper on the arXiv by Hanson et al. from SPTpol, an experiment which aims to measure the polarization of the cosmic microwave background using the South Pole Telescope. One of the main aims of experiments such as this is to measure the so-called “B-mode” of polarization (the “curl” component of the polarization signal, which possesses a handedness) because this holds the key to direct detection of a number of interesting cosmological phenomena such as the existence of primordial gravitational waves.  However, primordial effects aren’t  the only way to generate B-mode polarization. Other “foreground” effects can do the job too, especially gravitational lensing can also generate a signal of this form. These “late-time” effects have to be understood before the primordial contribution can be isolated.

Before today there was no direct measurement of B-mode polarization at all, primordial nor not.

The abstract basically says it all:

Gravitational lensing of the cosmic microwave background generates a curl pattern in the observed polarization. This “B-mode” signal provides a measure of the projected mass distribution over the entire observable Universe and also acts as a contaminant for the measurement of primordial gravity-wave signals. In this letter we present the first detection of gravitational lensing B modes, using first-season data from the polarization-sensitive receiver on the South Pole Telescope (SPTpol). We construct a template for the lensing B-mode signal by combining E-mode polarization measured by SPTpol with estimates of the lensing potential from a Herschel-SPIRE map of the cosmic infrared background. We compare this template to the B modes measured directly by SPTpol, finding a non-zero correlation at 7.7 sigma significance. The correlation has an amplitude and scale-dependence consistent with theoretical expectations, is robust with respect to analysis choices, and constitutes the first measurement of a powerful cosmological observable.

This measurement is not unexpected. Indeed, the B-mode contribution from lensing by the known distribution of galaxies can be calculated fairly straightforwardly because the physics is well understood; failure to find the expected signal would therefore have been somewhat embarrassing.  It’s a different story for the primordial B-mode because that depends strongly on what is going on in the very early universe, and that is much less certain. Although the new result doesn’t itself tell us anything new about the very early Universe it is definitely an important step on the way, and it’s a fairly safe prediction that there will be a great deal of activity and interest in CMB polarization over the next few years, including next year’s planned release of polarization data from Planck.

I’ll also note the use of Herschel-SPIRE images in tracing the galaxy images, in deference to my former colleagues in Cardiff who played a key role in developing that instrument!