Cosmic birefringence tomography and calibration independence with reionization signals in the CMB

The search for cosmic polarization rotation or birefringence in the cosmic microwave background (CMB) is well motivated because it can provide powerful constraints on parity-violating new physics, such as axion-like particles. In this paper, we point out that since the CMB polarization is produced at two very different redshifts - it is generated at both reionization and recombination - new parity-violating physics can generically rotate the polarization signals from these different sources by different amounts. We explore two implications of this. First, measurements of CMB birefringence are challenging because the effect is degenerate with a miscalibration of CMB polarization angles; however, by taking the difference of the reionization and recombination birefringence angles (measured from different CMB angular scales), we can obtain a cosmological signal that is immune to instrumental angle miscalibration. Secondly, we note that the combination with other methods for probing birefringence can give tomographic information, constraining the redshift origin of any physics producing birefringence. We forecast that the difference of the reionization and recombination birefringence angles can be competitively determined to within similar to 0.05 deg for future CMB satellites such as LiteBIRD. Although much further work is needed, we argue that foreground mitigation for this measurement should be less challenging than for inflationary B-mode searches on similar scales due to larger signals and lower foregrounds.

Automated summary

The search for cosmic polarization rotation or birefringence in the CMB is important for constraining new physics. This paper points out that new parity-violating physics can rotation the polarization signals from different redshifts, and proposes two implications: using the difference of birefringence angles to obtain a cosmological signal immune to miscalibration, and using other methods to constrain the redshift origin of birefringence. Future CMB satellites like LiteBIRD can competitively determine the difference of birefringence angles to within approximately 0.05 deg.