Cosmic birefringence​
​
​Cosmic birefringence refers to a possible rotation of the polarization plane of CMB photons as they propagate across cosmological distances (see Komatsu, Nature Reviews Physics 2022, for a review). Such a rotation can arise from parity-violating physics beyond the standard model, for example through coupling of photons to a dynamical pseudo-scalar field such as axion-like particles, a compelling dark matter candidate. This effect converts CMB E-modes into B-modes and generates TB and EB correlations that are otherwise forbidden in the standard cosmological model. Measuring or constraining cosmic birefringence is therefore of great interest as a probe of fundamental physics. However, similar EB cross-correlation signals can also be produced by instrumental miscalibration of detector polarization angles, making careful calibration and tailored component separation techniques essential for distinguishing between genuine cosmic birefringence and systematic effects. Unlike the CMB, Galactic foreground polarization can only be rotated by instrumental miscalibration, since their photons have not traveled across cosmological distances, providing a path to break the degeneracy between cosmic birefringence and systematics in the EB cross-power spectrum (Minami & Komatsu, PTEP 2020). Using this approach, preliminary evidence for a cosmic birefringence angle of β ≃ 0.3° ± 0.11° has been obtained from Planck data (Diego-Palazuelos et al., PRL 2022).
​​​​
(Click on paper's title to access the corresponding publication)
Remazeilles, arXiv:2507.22109 (2025)​​​​
​​​
In this paper, I proposed a new real-space, field-level methodology to infer cosmic birefringence directly from CMB fields rather than from the EB cross-power spectrum. I introduced a novel hybrid internal linear combination (ILC) approach that jointly combines E- and B-mode frequency maps and enables disentangling the correlated and uncorrelated components of the CMB polarization field. I then derived an analytic linear relation between the birefringence-induced correlated component of the CMB E-mode field and the full CMB B-mode field, allowing linear regression across sky patches to extract the birefringence angle. This approach, blind to prior foreground models, robustly distinguishes true cosmic birefringence from instrumental polarization angle miscalibration and foregrounds, as the ILC weights differently project achromatic versus chromatic effects. In LiteBIRD mission simulations, it achieves constraints that are competitive with and complementary to traditional power-spectrum approaches. Applied to Planck PR4 data, I obtained a 2.7σ detection of a birefringence angle of β = 0.32° ± 0.12°, which is furthermore consistent across varying sky fractions. This field-level method may open new opportunities to probe fundamental physics with the CMB.
​​