Exploring cosmic origins with CORE:B-mode component separation

M. Remazeilles, A. J. Banday, C. Baccigalupi, S. Basak, A. Bonaldi, G. de Zotti, C. L. Dickinson, H. K. Eriksen, Josquin Errard, R. Fernández-Cobos, U. Fuskeland, Carlos Hervías-Caimapo, M. López-Caniego, E. Martínez-González, M. Roman, P. Vielva, I. K. Wehus, A. Achúcarro, P. A. R. Ade, R. Allison, M. Ashdown, M. Ballardini, R. Banerji, J. G. Bartlett, N. Bartolo, Daniel Baumann, M. Bersanelli, Matteo Bonato, J. Borrill, F. R. Bouchet, F. Boulanger, T. Brinckmann, M. Bucher, C. Burigana, A. Buzzelli, Zhen-Yi Cai, M. Calvo Gomez, C. S. Carvalho, G. Castellano, A. Challinor, Jens Chluba, Sébastien Clesse, I. Colantoni, A. Coppolecchia, M. Crook, G. D’Alessandro, P. de Bernardis, G. de Gasperis, J. M. Diego, Eleonora Di Valentino, S. Feeney, Simone Ferraro, F. Finelli⋆, F. Forastieri, S. Galli, R. T. Génova-Santos, M. Gerbino, J. González-Nuevo, S. Grandis, J. Greenslade, Yilun Guan, Shaul Hanany, Will Handley, C. Hernández-Monteagudo, M. Hills, E. Hivon, K. Kiiveri, T. S. Kisner, T. Kitching, M. Kunz, H. Kurki‐Suonio, L. Lamagna, A. Lasenby, M. Lattanzi, J. Lesgourgues, Antony Lewis, M. Liguori, V. Lindholm, G. Luzzi, B. Maffei, C. J. A. P. Martins, S. Masi, S. Matarrese, D. McCarthy, J.‐b. Melin, A. Melchiorri, D. Molinari, A. Monfardini, S Nesseris, M. Negrello, Alessio Notari, A. Paiella, D. Paoletti, G. Patanchon, M. Piat, G. Pisano, L. Polastri, G. Polenta, A. PolloJournal of Cosmology and Astroparticle Physics 2018 · DOI

TL;DR

This paper demonstrates that the proposed CORE satellite mission can effectively separate the faint primordial CMB B-mode signal from overwhelming foregrounds, enabling the detection of primordial gravitational waves (characterized by the tensor-to-scalar ratio, r) down to r = 5 x 10⁻³. This is crucial for probing the physics of the early universe and the inflationary epoch.

What Problem Does This Paper Address?

The Cosmic Microwave Background (CMB) is a powerful probe of the early universe. Its polarization contains information about primordial gravitational waves generated during inflation, which leave a specific signature in the B-mode polarization. However, this primordial B-mode signal is incredibly faint and is completely swamped by much stronger foreground emissions from our own galaxy (synchrotron, dust, anomalous microwave emission) and extragalactic sources (radio and infrared). Detecting the primordial B-modes requires sophisticated component separation techniques to isolate the faint CMB signal from these astrophysical contaminants. This paper addresses the challenge of achieving the necessary foreground cleaning with the proposed CORE mission to detect primordial B-modes at scientifically interesting levels.

Methods

The authors employ three distinct component separation methods:

  • Commander: A Bayesian parametric fitting approach that models foregrounds with specific spectral laws and fits their parameters to the data. This method is computationally intensive and best suited for large angular scales.
  • NILC (Needlet Internal Linear Combination): A blind component separation method that uses needlets (wavelets localized in both real and harmonic space) to filter the sky and combine multi-frequency data to minimize foreground residuals. This method is applied at high multipoles.
  • SMICA (Spectral Matching Independent Component Analysis): Another blind method that jointly estimates spatial power spectra of components and noise by matching the data to a spectral model.

These methods are applied to simulated CORE sky maps that include realistic models of CMB, synchrotron, thermal dust, AME, and point sources, as well as gravitational lensing effects. The performance is evaluated by estimating the tensor-to-scalar ratio, r, and its uncertainty.

Datasets

The paper utilizes simulated sky maps generated using a modified version of the Planck Sky Model (PSM version 1.7.8). These simulations are designed to mimic the expected observations of the CORE satellite mission and include various combinations of astrophysical foregrounds and gravitational lensing effects. No real observational datasets are used for training or evaluation in this paper; the focus is on assessing the mission's capabilities through simulated data.

Key Insights and Ideas

  • Foregrounds are the primary challenge: The paper consistently highlights that even with advanced component separation techniques, foreground residuals remain the dominant source of uncertainty for detecting primordial B-modes, especially at lower r values.
  • Delensing helps, but doesn't solve everything: While delensing (removing the B-mode signal induced by gravitational lensing) reduces cosmic variance, it doesn't fully overcome the limitations imposed by foregrounds at low r.
  • Frequency coverage is critical: The authors emphasize that a wide frequency range, as provided by CORE, is essential for disentangling foregrounds, particularly to act as "lever arms" for modeling complex spectral behaviors.
  • Pixelization can introduce biases: The paper points out that the process of degrading sky maps to lower resolution can artificially introduce curvature into foreground spectral indices, potentially biasing component separation results.
  • Synergy between methods is beneficial: Combining the strengths of different component separation methods (e.g., Commander for low multipoles and SMICA/NILC for high multipoles) leads to more robust results.

Key Results

  • For r = 10⁻² and r = 5 × 10⁻³, CORE can achieve high-significance detections of primordial B-modes (exceeding 10σ) even with foregrounds and lensing.
  • At the critical level of r = 10⁻³, the detection becomes significantly more challenging. Without lensing, Commander alone can achieve an ~8σ detection.
  • With lensing and complex foregrounds (synchrotron, dust, AME, point sources), the uncertainty on r at the 10⁻³ level is around σ(r) ~ 0.5 × 10⁻³.
  • After accounting for foreground residuals and 60% delensing, CORE can constrain r to σ(r) ~ 3 × 10⁻⁴, which is a significant improvement and crucial for probing inflationary models.
  • The paper demonstrates that even with perfect foreground cleaning, lensing contributes to cosmic variance, but foreground residuals remain the dominant error source for r = 10⁻³.

Limitations and Future Work

  • Foreground modeling: The accuracy of foreground modeling is a persistent limitation. The paper suggests that external surveys (like C-BASS and QUIJOTE) and improved modeling of AME polarization will be crucial for future improvements.
  • Computational cost: The Commander algorithm's computational intensity limits its application to low-resolution maps. Future work should focus on optimizing it for higher resolutions.
  • Systematic errors: The paper acknowledges that systematic effects like calibration errors, beam asymmetries, and bandpass mismatch can impact component separation and foreground removal. These need to be addressed in future, more realistic simulations.
  • Delensing: While delensing is shown to be beneficial, the paper suggests that further optimization of delensing techniques and incorporating lensing potential reconstruction will be important.
  • Pixelization effects: The impact of pixelization on foreground spectral indices needs to be further investigated and mitigated in future analyses.

Summary updated

Abstract

The research leading to these results has received funding from the ERC Starting Consolidator Grant (no. 307209). Some of the results in this paper have been derived using the HEALPix package. We acknowledge the use of the PSM package, developed by the Planck working group on component separation, for making the simulations used in this work. We acknowledge the use of the Ulysses cluster at SISSA. This research was partially supported by the RADIOFOREGROUNDS project, funded by the European Commission’s H2020 Research Infrastructures under the Grant Agreement 687312, and the INDARK INFN Initiative. GDZ acknowledges support by ASI/INAF agreement no. 2014-024-R.1. R.F.-C., E.M.-G., and P.V. acknowledge support from the Spanish MINECO project ESP2015-70646-C2-1-R (cofinanced with EU FEDER funds), Consolider-Ingenio 2010 project CSD2010-00064 and from the CSIC “Proyecto Intramural Especial” project 201550E091. JGN acknowledges financial support from the Spanish MINECO for a ‘Ramon y Cajal’ fellowship (RYC-2013-13256) and the I+D 2015 project AYA2015-65887-P (MINECO/FEDER). CJM is supported by an FCT Research Professorship, contract reference IF/00064/2012, funded by FCT/MCTES (Portugal) and POPH/FSE.
BibTeX
@article{remazeilles2018exploring,
  author = {M. Remazeilles and A. J. Banday and C. Baccigalupi and S. Basak and A. Bonaldi and G. de Zotti and G. de Zotti and C. L. Dickinson and H. K. Eriksen and Josquin Errard and R. Fernández-Cobos and U. Fuskeland and Carlos Hervías-Caimapo and M. López-Caniego and E. Martínez-González and M. Roman and P. Vielva and I. K. Wehus and A. Achúcarro and P. A. R. Ade and R. Allison and M. Ashdown and M. Ballardini and R. Banerji and J. G. Bartlett and N. Bartolo and Daniel Baumann and M. Bersanelli and Matteo Bonato and J. Borrill and F. R. Bouchet and F. Boulanger and T. Brinckmann and M. Bucher and C. Burigana and A. Buzzelli and Zhen-Yi Cai and M. Calvo Gomez and C. S. Carvalho and G. Castellano and A. Challinor and Jens Chluba and Sébastien Clesse and I. Colantoni and A. Coppolecchia and M. Crook and G. D’Alessandro and P. de Bernardis and G. de Gasperis and J. M. Diego and Eleonora Di Valentino and S. Feeney and Simone Ferraro and F. Finelli⋆ and F. Forastieri and S. Galli and R. T. Génova-Santos and M. Gerbino and J. González-Nuevo and S. Grandis and J. Greenslade and Yilun Guan and Shaul Hanany and Will Handley and C. Hernández-Monteagudo and M. Hills and E. Hivon and K. Kiiveri and T. S. Kisner and T. Kitching and M. Kunz and H. Kurki‐Suonio and L. Lamagna and A. Lasenby and M. Lattanzi and J. Lesgourgues and Antony Lewis and M. Liguori and V. Lindholm and G. Luzzi and B. Maffei and C. J. A. P. Martins and S. Masi and S. Matarrese and D. McCarthy and J.‐B. Melin and A. Melchiorri and D. Molinari and A. Monfardini and S Nesseris and M. Negrello and Alessio Notari and A. Paiella and D. Paoletti and G. Patanchon and M. Piat and G. Pisano and L. Polastri and G. Polenta and A. Pollo},
  title = {{Exploring cosmic origins with CORE:<i>B</i>-mode component separation}},
  journal = {Journal of Cosmology and Astroparticle Physics},
  year = {2018},
  doi = {10.1088/1475-7516/2018/04/023},
}
Figure