Calibration of CMB Polarisation Using Cross-Experiment Correlations

arXiv:2606.11300 · astro-ph.CO, astro-ph.IM · Submitted 2026-06-09 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: We're looking at the paper: "Calibration of CMB Polarisation Using Cross-Experiment Correlations".

Jocelyn: Here's the gist of the paper. Calibration of CMB Polarisation Using Cross-Experiment Correlations describes a data-driven method to calibrate the relative polarisation angle between CMB experiments using cross-correlations of observations over a common sky region Unlike standard self-calibration approaches, this method does not assume vanishing isotropic cosmic birefringence or primordial EB correlations when estimating the relative misalignment angle, and therefore preserves sensitivity to parity-violating physics.

Vera: As a proof of concept, it forecasts the performance of this method using the Simons Observatory (SO) Small Aperture Telescopes (SATs) as a calibrated reference.

Jocelyn: The Problem of Miscalibration Accurate recovery of the CMB E- and B-mode signals requires precise knowledge of the instrument’s polarisation orientation.

Vera: A small miscalibration in the polarimeter orientation rotates the measured polarisation basis and induces leakage from E- to B-modes.

Title and authors: Vera: Okay, let’s talk about who wrote this and what they’re calling this work, "Calibration of CMB Polarisation Using Cross-Experiment Correlations." The authors are Claire Rigouzzo, Eugene Lim, Susanna Azzoni, and Yiqi Liu.

Jocelyn: It’s interesting that the title says it uses cross-experiment correlations to calibrate polarization. It sounds very practical because it’s about using data from different telescopes to check each other.

Subrahmanyan: The authors are tackling the fundamental problem of instrument calibration, which is usually a nuisance but becomes a serious cosmological constraint when you’re looking for subtle effects like cosmic birefringence or primordial electron-b correlations.

Vera: They introduce this data-driven method specifically because it avoids the pitfalls of standard self-calibration approaches by not assuming vanishing isotropic cosmic birefringence or primordial electron-b correlations when they estimate the relative misalignment angle.

Jocelyn: That’s a big deal, because as we saw on page three, that degeneracy between instrumental misalignment and cosmic birefringence significantly limits our ability to constrain the latter.

Subrahmanyan: So if this method can isolate the misalignment angle without being completely blind to those parity-violating signals, it opens up a different way to probe physics beyond the Standard Model.

The paper's summary: Vera: So, what’s the actual mechanism they propose? They use cross-correlations between different instruments to build an estimator for the relative misalignment angle. They call this quantity R sub i j, which directly estimates the difference between telescope misalignment angles using only their cross-spectra.

Jocelyn: That sounds like they are using the way one instrument’s measurement correlates with another one’s, even if they aren't observing exactly the same patch of sky simultaneously.

Subrahmanyan: The key thing about R sub i j is that it provides a direct estimate of that difference without needing any prior assumptions about the intrinsic electron-b mode spectrum or whether cosmic birefringence beta is present at all. That’s a very clean probe.

Vera: Right, because the paper says this estimator targets precisely the observable that is both measurable and free from degeneracy with parity-violating physics. It’s a very clever target for calibration.

Jocelyn: If you can measure something that isn't confused with the physics you’re trying to find, that makes it much more useful for constraining things like primordial electron-b correlations.

The paper's improvements: Vera: The practical improvement they highlight is how this method can be used to improve constraints on parity-breaking physics in Section IV of the paper. They show how you can assess whether parity-violating physics is present by looking at the case where beta equals zero, which helps constrain the primordial electron-b mode spectrum.

Jocelyn: But they also showed that assuming beta equals zero and using only one instrument lets you read out the primordial electron-b mode power spectrum. That comparison really shows you how much precision you can get depending on what data set you use.

Subrahmanyan: What’s interesting is the performance forecast they give, using the Simons Observatory Small Aperture Telescopes, or SATs, as a calibrated reference point. They show that if those SATs can be calibrated to an uncertainty of zero point zero eight degrees, then Planck and LAT could potentially be calibrated to uncertainties of zero point one zero degrees and zero point one seven degrees at around one hundred forty-five gigahertz.

Vera: So the forecast shows a real potential for retrospective cross-calibration. It suggests that existing datasets from experiments like Planck and LAT can be tightened significantly if they use this method against a well-calibrated reference like the SATs.

Conclusion: Jocelyn: So, to wrap up, the central result of this paper is that an antisymmetric combination of cross-instrument electron-b mode spectra isolates this relative misalignment angle while remaining insensitive to both cosmic birefringence and intrinsic electron-b signals. It’s a robust way to extract what you need.

Subrahmanyan: This approach enables a significant reduction in systematic uncertainties, which means we can get better constraints on parity-violating physics in the early universe. It really shows the power of cross-instrument analyses for improving polarization calibration across different experiments.

Vera: The key result is that this antisymmetric combination of cross-instrument electron-b mode spectra isolates this relative angle while remaining insensitive to both birefringence and intrinsic electron-b signals. That’s what makes it so powerful, because it removes the noise from those other physical effects.

Jocelyn: It identifies a specific observable that can be robustly extracted from two experiments at the level of two-point functions and provides a direct estimator for it. The forecasts show that assuming a well-calibrated reference instrument such as the Simons Observatory SATs, this method can achieve polarization-angle uncertainties of zero point one zero degrees for the LAT and zero point one seven degrees for Planck at one hundred forty-five gigahertz including both systematic and statistical errors.

Subrahmanyan: And while they show how to get better calibration, they also flag a limitation: the largest source of uncertainty comes from the pixel covariance term, which tells us that even with this method, you still have some room for error in achieving that sharp precision.

Vera: So it’s a great tool for improving calibration and probing new physics at the level of two-point functions, but we need to be careful about those covariance terms when we try to achieve those tight error bars. That’s what this paper on "Calibration of CMB Polarisation Using Cross-Experiment Correlations" shows us.

Laboratory for Theoretical Particle Physics and Cosmology, King’s College London · Joseph Henry Laboratories of Physics, Princeton University

astro-ph.CO, astro-ph.IM

Submitted: 2026-06-09

Updated: 2026-11-10

Comments: 12 pages, 4 figures, 3 appendices. Published in the Open Journal of Astrophysics

Code: https://github.com/simonsobs/so_noise_models

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 91/100

The gist: Calibration of CMB Polarisation Using Cross-Experiment Correlations describes a data-driven method to calibrate the relative polarisation angle between CMB experiments using cross-correlations of…

Key concepts

Miscalibration Problem
When instrument polarimeters are misaligned, they rotate the measured polarization basis. This rotation causes leakage of E-mode power into B-mode measurements, complicating the extraction of true sky signals and parity-violating physics.
Cross-Correlation Estimator (Rij)
This estimator uses cross-correlations between independent instruments to directly estimate the difference in their telescope misalignment angles. Crucially, it does not require assumptions about the intrinsic CMB spectrum or cosmic birefringence, making it model-independent for this specific parameter.
Parity Violation Probe
The method can test for parity violation by setting the cosmic birefringence angle ($eta$) to zero and measuring the primordial C_EB power spectrum. Conversely, using only one instrument with $eta=0$ allows reading the primordial C_EB spectrum, showing how this technique separates calibration errors from fundamental physics signals.
Reference Instrument Calibration
The performance of the method is forecasted using a well-calibrated reference, such as the Simons Observatory SATs. This demonstrates that by anchoring relative measurements to a high-precision instrument, existing datasets like Planck and LAT can be calibrated with improved accuracy.

Terminology

Summary

Calibration of CMB Polarisation Using Cross-Experiment Correlations describes a data-driven method to calibrate the relative polarisation angle between CMB experiments using cross-correlations of observations over a common sky region Unlike standard self-calibration approaches, this method does not assume vanishing isotropic cosmic birefringence or primordial EB correlations when estimating the relative misalignment angle, and therefore preserves sensitivity to parity-violating physics. p. 0

Where the paper says it

We describe a data-driven method to calibrate the relative polarisation angle between CMB experiments using cross-correlations of observations over a common sky region. Unlike standard self-calibration approaches, this method does not assume vanishing isotropic cosmic birefringence or primordial EB correlations when estimating the relative misalignment angle, and therefore preserves sensitivity to parity-violating physics.

As a proof of concept, it forecasts the performance of this method using the Simons Observatory (SO) Small Aperture Telescopes (SATs) as a calibrated reference. p. 0
Where the paper says it

Unlike standard self-calibration approaches, this method does not assume vanishing isotropic cosmic birefringence or primordial EB correlations when estimating the relative misalignment angle, and therefore preserves sensitivity to parity-violating physics. As a proof of concept, we forecast the performance of this method using the Simons Observatory (SO) Small Aperture Telescopes (SATs) as a calibrated reference.

The Problem of Miscalibration

Accurate recovery of the CMB E- and B-mode signals requires precise knowledge of the instrument’s polarisation orientation. A small miscalibration in the polarimeter orientation rotates the measured polarisation basis and induces leakage from E- to B-modes. p. 1

Where the paper says it

THE PROBLEM OF MISCALIBRATION Accurate recovery of the CMB E- and B-mode signals requires precise knowledge of the instrument’s polarisation orientation. A small miscalibration in the polarimeter orientation rotates the measured polarisation basis and induces leakage from E- to B-modes [17–19].

The observed EB power spectrum thus receives two contributions: a sky term C EB,l, and a rotation-induced term proportional to C EE,l − C BB,l cos(2α) sin(2α). p. 1
Where the paper says it

The observed EB power spectrum therefore reads [41, 44]: C EB,o ℓ = C EB ℓ cos(4α) + C EE ℓ − C BB ℓ cos(2α) sin(2α) . (3) The observed C EB,o ℓ thus receives two contributions: a sky term C EB ℓ , and a rotation-induced term proportional to C EE ℓ − C BB ℓ .

The goal is to disentangle the contribution arising from the misalignment angle from the true sky signal C EB,l, which encodes information about parity violation. p. 1
Where the paper says it

, cosmic birefringence) or astrophysical foregrounds – and instrumental polarisation angle miscalibration. Our goal is to disentangle the contribution arising from the misalignment angle from the true sky signal C EB ℓ , which encodes information about parity violation.

How the training signal is built

The method exploits cross-correlations between different instruments to construct an estimator of the relative misalignment angle by exploiting cross-correlations between independent instruments. The quantity Rij provides a direct estimate of the difference between the telescope misalignment angles using only their cross-spectra, without requiring prior assumptions about the intrinsic C EB,l spectrum or the presence of cosmic birefringence β. p. 3

Where the paper says it

| {z } ≡Rij . (8) The ratio Rij provides a direct estimate of the difference between the telescope misalignment angles using only their cross-spectra, without requiring prior assumptions about the intrinsic C EB ℓ spectrum or the presence of cosmic birefringence β.

This makes the estimator particularly powerful: it targets precisely the observable that is both measurable and free from degeneracy with parity-violating physics. p. 4
Where the paper says it

5 particularly powerful: it targets precisely the observable that is both measurable and free from degeneracy with parity-violating physics. It therefore provides a modelindependent probe of the relative polarisation-angle calibration.

Where the gains come from

The method's performance is forecasted using the Simons Observatory (SO) Small Aperture Telescopes (SATs) as a calibrated reference. The resulting calibration accuracy can also be leveraged by the LAT and other CMB experiments observing overlapping regions of the sky. p. 1

Where the paper says it

2 campaigns. We will show that the resulting calibration accuracy can also be leveraged by the LAT and other CMB experiments observing overlapping regions of the sky.

For instance, assuming a SAT calibration uncertainty of 0. p. 7
Where the paper says it

In the case of the polarised Galactic foreground calibration method, for instance, incorporating all frequency channels led to a twofold reduction in the error, see [42] in comparison with [68]. A more detailed implementation of this SAT–LAT crosscalibration approach, including map-based simulations, an exploration of different calibration regimes, updated SO noise models, foreground treatment, will be presented in an upcoming Simons Observatory paper [47].

08◦, the LAT/ Planck can be calibrated to uncertainties of 0. p. 0
Where the paper says it

If they can be calibrated to an uncertainty of 0. 08◦ , as anticipated from the SO wire grid calibration system, we show that the SO Large Aperture Telescope and Planck could be calibrated to uncertainties of 0.

10◦ and 0.17◦ at ∼ 145 GHz. p. 0
Where the paper says it

10◦ and 0. 17◦ , respectively, at ∼ 145 GHz.

Constraining Parity-Violation

The method can be used to assess whether parity-violating physics is present by separately investigating the case where β = 0, to constrain the primordial C EB,l. Conversely, assuming β = 0 and using only one instrument allows reading the primordial C EB,l power spectrum [42]. p. 6

Where the paper says it

Using only one instrument, from Eq. (3), we can read the primordial C EB ℓ power spectrum [42]: C EB ℓ ≃ C EB,o ℓ,ii − 2αi C EE,o ℓ,ii − C BB,o ℓ,ii .

This demonstrates that the SAT is achieving the best precision for constraining the primordial C EB,l signal, but has a very limited l range. p. 6
Where the paper says it

8 This gives us an idea of the amplitude and shape necessary to be detectable with our current calibration method. 4 we observe that the SAT is achieving the best precision, but has a very limited ℓ range.

Discussion and Outlook

The central result is that an antisymmetric combination of cross-instrument EB spectra isolates this relative angle while remaining insensitive to both birefringence and intrinsic EB signals. p. 7

Where the paper says it

By leveraging cross-correlations between different experiments [30], we demonstrated that any datasets that share some sky coverage can be recalibrated using the SO SATs. The central result is that an antisymmetric combination of cross-instrument EB spectra isolates this relative angle while remaining insensitive to both birefringence and intrinsic EB signals, as shown in Eq.

This approach enables a significant reduction of systematic uncertainties and enhances our ability to constrain or detect signatures of parity-violating physics in the early universe. p. 8
Where the paper says it

Encouragingly, additional calibration strategies, including drone-based systems, are being developed and offer promising avenues for achieving highprecision absolute calibration. This approach enables a significant reduction of systematic uncertainties and enhances our ability to constrain or detect signatures of parity-violating physics in the early universe.

The method highlights the potential of cross-instrument analyses to improve polarisation calibration across experiments. p. 8
Where the paper says it

This approach enables a significant reduction of systematic uncertainties and enhances our ability to constrain or detect signatures of parity-violating physics in the early universe. More generally, this method highlights the potential of cross-instrument analyses to improve polarisation calibration across experiments.

Conclusion

In this work, we have examined the impact of polarisation-angle miscalibration as a key limitation in extracting parity-violating signals from the CMB. The central result is that an antisymmetric combination of cross-instrument EB spectra isolates this relative angle while remaining insensitive to both birefringence and intrinsic EB signals. This identifies the specific observable that can be robustly extracted from two experiments at the level of two-point functions and provides a direct estimator for it. p. 7

Where the paper says it

The central result is that an antisymmetric combination of cross-instrument EB spectra isolates this relative angle while remaining insensitive to both birefringence and intrinsic EB signals, as shown in Eq. This identifies the specific observable that can be robustly extracted from two experiments at the level of two-point functions, and provides a direct estimator for it.

The forecasts show that, assuming a well-calibrated reference instrument such as the Simons Observatory SATs, this method can achieve polarisation-angle uncertainties of 0. p. 7
Where the paper says it

While cross-correlations are widely used in CMB analyses, this formulation makes explicit how they can be used to calibrate polarisation angles without imposing assumptions that would otherwise remove the very signals of interest. Our forecasts show that, assuming a well-calibrated reference instrument such as the Simons Observatory SATs, this method can achieve polarisation-angle uncertainties of 0.

10◦ for the LAT and 0. p. 0
Where the paper says it

As a concrete application, we demonstrate how this approach can leverage the expected precise calibration of the SO SATs [31–34] to calibrate other telescopes. The Simons Observatory, located in Chile, consists of two complementary surveys using the SATs and the Large Aperture Telescope (LAT) [35].

17◦ for Planck at 145/143 GHz (including both systematic and statistical errors). p. 7
Where the paper says it

10◦ for the LAT and 0. 17◦ for Planck at 145/143 GHz (including both systematic and statistical errors).

The key requirement of this approach is the availability of at least one accurately calibrated instrument to anchor the relative measurements. p. 7
Where the paper says it

This demonstrates that cross-instrument calibration can significantly improve the effective calibration of existing datasets, thereby enhancing their sensitivity to parity-violating physics. A key requirement of this approach is the availability of at least one accurately calibrated instrument to anchor the relative measurements.

Acknowledgments

We are very grateful to Kam Arnold, Carlos Herv´ıas-Caimapo, Jo Dunkley, Lam Hui, Mudit Jain, Anto Lonappan and David Marsh for discussions and insightful feedback. We would also like to thank the organisers of the GC2024 conference at the Yukawa Institute where this work was originally conceived in a discussion between the authors and Jo Dunkley. p. 8

Where the paper says it

Acknowledgments: We are very grateful to Kam Arnold, Carlos Herv´ıas-Caimapo, Jo Dunkley, Lam Hui, Mudit Jain, Anto Lonappan and David Marsh for discussions and insightful feedback. We would also like to thank the organisers of the GC2024 conference at the Yukawa Institute where this work was originally conceived in a discussion between the authors and Jo Dunkley.

C.R. acknowledges support from the Science and Technology Facilities Council (STFC). p. 8
Where the paper says it

We would also like to thank the organisers of the GC2024 conference at the Yukawa Institute where this work was originally conceived in a discussion between the authors and Jo Dunkley. acknowledges support from the Science and Technology Facilities Council (STFC).

Appendix A: Isotropic cosmic birefringence

We demonstrate how the presence of parity-violating interactions affecting photons can be probed through the rotation of their polarisation plane. Consider a parity violating pseudoscalar field ϕ, that couples to photons through a Chern-Simons term ϕFµνF˜µν, where F˜µν = 1/2ϵµναβFαβ. p. 1

Where the paper says it

Isotropic Cosmic Birefringence The presence of parity-violating interactions for photons can be measured via the rotation of their polarisation plane. Consider for example a pseudoscalar field ϕ, that couples to photons through a Chern-Simons term ϕFµνF˜µν, where F˜ µν = 1/2ϵµναβF αβ.

As is shown in Appendix A, the presence of such a term would impact the right-handed and left-handed photons asymmetrically, rotating the plane of polarisation between the last scattering surface and today. p. 1
Where the paper says it

Consider for example a pseudoscalar field ϕ, that couples to photons through a Chern-Simons term ϕFµνF˜µν, where F˜ µν = 1/2ϵµναβF αβ. As is shown in Appendix A, the presence of such a term would impact the right-handed and left-handed photons asymmetrically, rotating the plane of polarisation between the last scattering surface and today.

Appendix B: What we cannot do with cross-correlating between different detectors

Unfortunately there exists no such relationship for ˜α+ij – one cannot “undo” the rotation of R(2˜αi) by adding another rotation to it since we do not know what the original datum point is. In terms of correlation functions, this conundrum manifests itself as an identity as follows. p. 8

Where the paper says it

(B4) Again, ˜α + ij is not present. In terms of correlation functions, this conundrum manifests itself as an identity as follows.

Appendix C: Additional data

Table IV provides a comparison of the calibration precision σRij between analytical estimates and simulations for SAT + Planck and SAT + LAT frequency combinations. Table V compares the 1σ deviation on the isotropic birefringence angle β for LAT and Planck for different frequency channels. p. 9

Where the paper says it

48 Table V. Comparison of the 1σ deviation on the isotropic birefringence angle β for LAT and Planck for different frequency channels.

The largest source of uncertainty comes from the pixel covariance term: ΞEB,o,l,ii, which indicates that the calibration uncertainty has reached sharp enough precision. p. 7
Where the paper says it

For these forecasts, we only use the 143/145, GHz channel. the largest source of uncertainty comes from the pixel covariance term: ΞEB,o ℓ,ii , which indicates that the calibration uncertainty has reached sharp enough precision.

p. 7
Where the paper says it

For these forecasts, we only use the 143/145, GHz channel. the largest source of uncertainty comes from the pixel covariance term: ΞEB,o ℓ,ii , which indicates that the calibration uncertainty has reached sharp enough precision.

Table III Summary of calibration methods

Type of calibration Precision C EB l = 0? β = 0? References. p. 3

Where the paper says it

4 Type of calibration Precision C EB ℓ = 0? References Self-calibration 0.

The wire-grid calibration is currently planned only for the SO SATs, and its achievable precision will be validated with additional on-sky data. p. 3
Where the paper says it

Summary of existing polarisation angle calibration methods, including their typical precision and the assumptions they require on the parity-violating sky signal. *Note that the wire-grid calibration is currently planned only for the SO SATs, and its achievable precision will be validated with additional on-sky data.

Table IV Comparison of calibration precision

Experiment pair f [GHz] σα [◦] σα˜ [◦] σβ [◦] References. The correlation SATs + LAT is 0. p. 1

Where the paper says it

We will show that the resulting calibration accuracy can also be leveraged by the LAT and other CMB experiments observing overlapping regions of the sky. We describe how this method can be used to calibrate other instruments via cross correlation with the SATs, specifically for Planck [39, 40] and LAT data.

10◦ and the correlation SATs + Planck is 0. p. 1
Where the paper says it

We will show that the resulting calibration accuracy can also be leveraged by the LAT and other CMB experiments observing overlapping regions of the sky. We describe how this method can be used to calibrate other instruments via cross correlation with the SATs, specifically for Planck [39, 40] and LAT data.

17◦.

Table V Comparison of the 1σ deviation on the isotropic birefringence angle β

LAT 93 0.17. Planck 143 0. p. 5

Where the paper says it

6 reference instrument and either the LAT or Planck as the second dataset. For the 143 GHz channel, we find: ( σR12 ≃ 0.

30. The largest source of uncertainty comes from the pixel covariance term: ΞEB,o,l,ii, which indicates that the calibration uncertainty has reached sharp enough precision.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Cross-instrument calibration for CMB experiments allows datasets to be recalibrated using a well-calibrated reference instrument, such as Simons Observatory SATs, enabling retrospective cross-calibration of existing datasets at minimal additional cost.

  2. The proposed method determines the relative misalignment angle between two instruments by exploiting cross-correlations between different instruments, which is robust to the presence of parity-violating signals such as cosmic birefringence or primordial C EB l.

  3. The estimator derived from Eq. (8) provides a modelindependent probe of the relative polarisation-angle calibration because it targets precisely the observable that is both measurable and free from degeneracy with parity-violating physics.

  4. Improved AI systems can perform precise estimation of the difference between telescope misalignment angles using only their cross-spectra, without requiring prior assumptions about intrinsic C EB l spectrum or the presence of cosmic birefringence beta.

  5. The method forecasts calibration uncertainties for Planck and LAT based on an assumed 0.08 degree precision for the SO SATs, showing that this approach can significantly improve the effective calibration of existing datasets.

Abstract

Parity-violating physics in the Universe can generate correlations between the Cosmic Microwave Background (CMB) E - and B-modes, but detecting such signals requires extremely accurate calibration of instruments. We describe a data-driven method to calibrate the relative polarisation angle between CMB experiments using cross-correlations of observations over a common sky region. Unlike standard self-calibration approaches, this method does not assume vanishing isotropic cosmic birefringence or primordial EB correlations when estimating the relative misalignment angle, and therefore preserves sensitivity to parity-violating physics. As a proof of concept, we forecast the performance of this method using the Simons Observatory (SO) Small Aperture Telescopes (SATs) as a calibrated reference. If they can be calibrated to an uncertainty of 0.08, as anticipated from the SO wire grid calibration system, we show that the SO Large Aperture Telescope and Planck could be calibrated to uncertainties of 0.10 and 0.17, respectively, at about 145 GHz. This approach relies on the availability of at least one well-calibrated instrument, and provides a complementary path to improving polarisation calibration across experiments, enabling more robust searches for parity-violating physics in the CMB, such as cosmic birefringence.

Sources

Related papers