BICEP2 I: Detection Of B-mode Polarization at Degree Angular Scales

TL;DR

BICEP2 reports a significant detection of B-mode polarization at degree angular scales, with a best-fit tensor-to-scalar ratio of $r=0.20^{+0.07}_{-0.05}$, suggesting a potential imprint of inflationary gravitational waves. The analysis employs advanced time-stream filtering, deprojection of beam systematics, and a matrix-based B-mode purification to isolate pure B modes, supported by extensive simulations and cross-spectra with BICEP1 and Keck Array data. Foreground analyses indicate dust and synchrotron are unlikely to fully account for the signal, though uncertainties remain, and the results show initial tension with temperature-based constraints, potentially resolvable with new physics or running. The work marks a pivotal step toward B-mode cosmology, illustrating both the potential detection of primordial tensor modes and the challenges posed by foregrounds and cross-validation across experiments.

Abstract

(abridged for arXiv) We report results from the BICEP2 experiment, a cosmic microwave background (CMB) polarimeter specifically designed to search for the signal of inflationary gravitational waves in the B-mode power spectrum around $\ell\sim80$. The telescope comprised a 26 cm aperture all-cold refracting optical system equipped with a focal plane of 512 antenna coupled transition edge sensor 150 GHz bolometers each with temperature sensitivity of $\approx300μ\mathrm{K}_\mathrm{CMB}\sqrt{s}$. BICEP2 observed from the South Pole for three seasons from 2010 to 2012. A low-foreground region of sky with an effective area of 380 square deg was observed to a depth of 87 nK deg in Stokes $Q$ and $U$. We find an excess of $B$-mode power over the base lensed-LCDM expectation in the range $30< \ell< 150$, inconsistent with the null hypothesis at a significance of $> 5σ$. Through jackknife tests and simulations we show that systematic contamination is much smaller than the observed excess. We also examine a number of available models of polarized dust emission and find that at their default parameter values they predict power $\sim(5-10)\times$ smaller than the observed excess signal. However, these models are not sufficiently constrained to exclude the possibility of dust emission bright enough to explain the entire excess signal. Cross correlating BICEP2 against 100 GHz maps from the BICEP1 experiment, the excess signal is confirmed and its spectral index is found to be consistent with that of the CMB, disfavoring dust at $1.7σ$. The observed $B$-mode power spectrum is well fit by a lensed-LCDM + tensor theoretical model with tensor-to-scalar ratio $r=0.20^{+0.07}_{-0.05}$, with $r=0$ disfavored at $7.0σ$. Accounting for the contribution of foreground dust will shift this value downward by an amount which will be better constrained with upcoming data sets.

Figures (14)

• Figure 1: BICEP2 $T$, $Q$, $U$ maps. The left column shows the basic signal maps with $0.25^\circ$ pixelization as output by the reduction pipeline. The right column shows difference (jackknife) maps made with the first and second halves of the data set. No additional filtering other than that imposed by the instrument beam (FWHM $0.5^\circ$) has been done. Note that the structure seen in the $Q$ and $U$ signal maps is as expected for an $E$-mode dominated sky.
• Figure 2: BICEP2 power spectrum results for signal (black points) and temporal-split jackknife (blue points). The solid red curves show the lensed-$\Lambda$CDM theory expectations while the dashed red curves show $r=0.2$ tensor spectra and the sum of both. The error bars are the standard deviations of the lensed-$\Lambda$CDM+noise simulations and hence contain no sample variance on tensors. The probability to exceed (PTE) the observed value of a simple $\chi^2$ statistic is given (as evaluated against the simulations). Note the very different $y$-axis scales for the jackknife spectra (other than $BB$). See the text for additional discussion of the $BB$ spectrum. (Note that the calibration procedure uses $EB$ to set the overall polarization angle so $TB$ and $EB$ as plotted above cannot be used to measure astrophysical polarization rotation---see Sec. \ref{['sec:polrot']}.)
• Figure 3: Left: BICEP2 apodized $E$-mode and $B$-mode maps filtered to $50<\ell<120$. Right: The equivalent maps for the first of the lensed-$\Lambda$CDM+noise simulations. The color scale displays the $E$-mode scalar and $B$-mode pseudoscalar patterns while the lines display the equivalent magnitude and orientation of linear polarization. Note that excess $B$ mode is detected over lensing+noise with high signal-to-noise ratio in the map ($s/n>2$ per map mode at $\ell\approx70$). (Also note that the $E$-mode and $B$-mode maps use different color and length scales.)
• Figure 4: Distributions of the jackknife $\chi^2$ and $\chi$ PTE values over the 14 tests and three spectra given in Table \ref{['tab:ptes']}. These distributions are consistent with uniform.
• Figure 5: Left: $BB$ spectra from $T$-only input simulations using the measured per channel beam shapes compared to the lensed-$\Lambda$CDM+$r=0.2$ spectrum. From top to bottom the curves are (i) no deprojection, (ii) deprojection of differential pointing only (dp), (iii) deprojection of differential pointing and differential gain of the detector pairs (dp+dg), (iv) adding deprojection of differential beam width (dp+dg+bw), and (v) differential pointing, differential gain, and differential ellipticity (dp+dg+ellip). Right: Estimated levels of other systematics as compared to the lensed-$\Lambda$CDM+$r=0.2$ spectrum. Solid lines indicate expected contamination. Dashed lines indicate upper limits. All systematics are comparable to or smaller than the extended beam mismatch upper limit.