Sunyaev-Zeldovich effects​
​
Galaxy clusters are the most massive gravitationally bound structures in the Universe. Their hot intracluster gas, reaching temperatures of 10 to 100 million degrees, contains energetic free electrons that scatter CMB photons via Compton scattering, producing a y-type spectral distortion of the CMB blackbody radiation at the location of galaxy clusters on the sky: this is known as the thermal Sunyaev-Zeldovich (SZ) effect (Sunyaev & Zeldovich, 1972). The thermal SZ (tSZ) effect is independent of redshift and has a distinctive spectral signature that enables the detection of the most distant galaxy clusters through multi-frequency observations, while also probing the physical properties of baryonic and dark matter—making it a powerful cosmological probe. In addition, the bulk motions of clusters relative to the cosmic rest frame induce Doppler shifts in the CMB, giving rise to the kinetic SZ (kSZ) effect. The relativistic nature of hot electrons also imprints subtle temperature-dependent spectral distortions onto the thermal SZ signal, referred to as the relativistic SZ (rSZ) effect. Together, these SZ signals provide unique insights into large-scale structure, cluster astrophysics, and fundamental cosmological parameters.
​​​​
(Click on paper's title to access the corresponding publication)
Remazeilles & Chluba, MNRAS 538, 1576 (2025)​​​​
​​
In this paper, we presented new evidence for the relativistic SZ (rSZ) effect in the publicly available Planck CMB maps, including those previously considered thermal SZ-free, such as the Planck SMICA-noSZ and Constrained ILC (CILC) CMB maps, by stacking them on galaxy cluster positions. We inferred the average temperature of Planck clusters from a new perspective by applying the non-parametric, map-based technique of Remazeilles & Chluba, MNRAS (2020) to stacked cluster samples. As part of this analysis, we extracted and mapped the first-order rSZ moment field, y(Tâ‚‘ − Tâ‚‘*), from Planck data using pivot temperatures Tâ‚‘*​ ranging from 2 to 10 keV. Stacking these maps on clusters, we found that the signal cancels out for a pivot of Tâ‚‘* = 5 keV, revealing that the average electron temperature of Planck clusters is Tâ‚‘ ≈ 5 keV . Consistently, we inferred Tâ‚‘ = 4.9 ± 2.6 keV from the ratio of the Planck y- and yTâ‚‘-map fluxes. Building on this finding, we constructed an rSZ-free Planck CMB map by deprojecting the full relativistic SZ contribution using CILC, assuming an rSZ spectrum with Tâ‚‘ = 5 keV. Unlike previous tSZ-free CMB maps, this rSZ-free Planck CMB map shows no residual cluster contamination after stacking.​​​​
​
LiteBIRD science goals and forecasts. Mapping the hot gas in the Universe
Remazeilles et al. (100+ co-authors), JCAP 12, 026 (2024)
​​In this LiteBIRD Collaboration paper, we demonstrated the LiteBIRD space mission's capability to map the thermal SZ effect, arising from the scattering of CMB photons by hot baryonic gas within and around galaxy clusters. LiteBIRD, although primarily designed to search for large-scale CMB B-mode polarization, will deliver full-sky maps in total intensity with unprecedented sensitivity across 15 frequency bands ranging from 40 to 402 GHz, thus offering an extensive dataset for detailed extraction and mapping of the thermal SZ effect. Despite its lower angular resolution, LiteBIRD outperforms ESA's Planck mission in both sensitivity and frequency coverage, allowing for the reconstruction of the next-generation all-sky map of the thermal SZ Compton y-parameter ("y-map") with significantly reduced foreground contamination at large and intermediate angular scales compared to Planck. Furthermore, we showed that combining LiteBIRD and Planck datasets within the component separation pipeline enables the production of an optimized LiteBIRD-Planck joint y-map that leverages LiteBIRD's sensitivity and Planck's angular resolution, improving cosmological constraints on σ₈ by about 15% compared to the Planck y-map. This paper also discusses the cosmological potential unlocked by a clean, all-sky map of diffuse thermal SZ emission from LiteBIRD.
​​​​
An improved Compton parameter map of thermal Sunyaev-Zeldovich effect from Planck PR4 data
Chandran, Remazeilles, Barreiro, MNRAS, 526, 5682 (2023)
​
In this paper, my PhD student Jyothis Chandran presented an improved all-sky Compton-y map of the thermal SZ effect constructed from the latest Planck PR4 data release. Using a customized Needlet Internal Linear Combination (NILC) pipeline applied to the nine Planck frequency channels, and validated on realistic sky simulations, we reconstructed the Planck SZ y-map over 98% of the sky with significantly improved quality compared to the previous public PR2 version (Planck 2015 results XXII, A&A 2016). The new PR4 NILC y-map shows a strong reduction of large-scale striping from by 1/f noise and reduced thermal dust contamination near the Galactic plane. At small angular scales, it also benefits from a ∼7% reduction in residual instrumental noise and a ∼34% reduction of CIB contamination compared to the PR2 version. These improvements are confirmed by angular power spectra, one-point statistical analyses, and visual inspections of the maps. As a result, the PR4 NILC y-map and associated products have been made publicly available on the Planck Legacy Archive, providing the community with a significantly cleaner all-sky y-map for astrophysical and cosmological studies.
​​​​
Mapping the relativistic electron gas temperature across the sky
Remazeilles & Chluba, MNRAS 494, 5734 (2020)
​​
In this paper, we introduced a novel component separation method to map the electron gas temperature (Tâ‚‘) of galaxy clusters over the entire sky using the relativistic SZ effect. Our approach combines the relativistic SZ temperature-moment expansion (Chluba et al., MNRAS 2013) with the Constrained ILC technique (Remazeilles et al., MNRAS 2011) to extract the zeroth- and first-order moment maps of the relativistic SZ emission, y and yTâ‚‘, which have distinct spectral signatures, from foreground-contaminated CMB data. Our method opens a new spectroscopic window on galaxy clusters, across both frequency and temperature. It also provides a new map-based observable: the electron temperature power spectrum, Tₑʸʸ(â„“), which complements the usual Compton-y power spectrum, Cʸʸ(â„“), through its different dependence on cosmological parameters, helping to break degeneracies between σ₈ and the hydrostatic mass bias. Accurate recovery of cluster temperatures across the sky, as demonstrated on LiteBIRD and PICO full-sky simulations, provides a powerful new proxy for cluster masses via the relativistic SZ effect and offers the possibility to disentangle electron density and temperature in cluster pressure profiles. Our technique has since been successfully applied to Planck data (Remazeilles & Chluba, MNRAS 2025) and ACT data (Coulton et al., 2025), while also demonstrating promising results for future SO data (Kuhn et al., 2025).
​​​​
Can we neglect relativistic temperature corrections in the Planck thermal SZ analysis?
Remazeilles, Bolliet, Rotti, Chluba, MNRAS 483, 3459 (2019)
​
All previous cluster cosmology studies—whether based on y-maps or cluster number counts—assumed the non-relativistic limit of the thermal SZ effect, neglecting relativistic corrections from the electron gas temperature (Tâ‚‘). In this paper, we proposed that including these relativistic corrections (the relativistic SZ effect) could help alleviate the tension in the cosmological parameter σ₈ between SZ clusters and the primary CMB, as originally reported in Planck papers. We revisited the construction of the Planck thermal SZ y-map by incorporating relativistic Tâ‚‘ corrections into the component separation pipeline, and recomputed the SZ power spectrum and skewness of the y-field for different electron temperatures. We demonstrated that both the SZ power spectrum amplitude and the skewness increase with Tâ‚‘, and estimated a resulting correction to σ₈ of Δσ₈/σ₈ ≃ 0.019 (kTâ‚‘ / 5 keV). For an average electron temperature of kTâ‚‘ ≳ 5 keV, this reduces the σ₈ tension between CMB and SZ clusters by more than 1σ.​​​
​
Planck 2015 results. XXII. A map of the thermal Sunyaev-Zeldovich effect
Planck Collaboration (co-corresponding author: Remazeilles; 200+ co-authors), A&A 594, A22 (2016)
​​
In this Planck Collaboration paper, I produced the first all-sky map of the thermal SZ effect, the Planck NILC y-map, publicly released on the Planck Legacy Archive. To achieve this, I implemented on Planck data the blind, wavelet-based component separation method NILC (Delabrouille et al., A&A 2009), specifically adapted for SZ reconstruction (Remazeilles et al., MNRAS 2011, 2013). The resulting map reveals galaxy clusters and diffuse hot gas across the entire sky through their Compton-y parameter. From the statistical properties of the SZ map (power spectrum and bispectrum), we obtained new constraints on the amplitude σ₈ of dark matter fluctuations, independent of those derived from the primary CMB. The observed tension between the σ₈ values inferred from CMB and SZ analyses has since motivated numerous studies worldwide, including scenarios involving massive neutrinos.
​​
Reconstruction of high-resolution Sunyaev-Zeldovich maps from heterogeneous data sets using needlets
Remazeilles, Aghanim, Douspis, MNRAS 430, 370 (2013)
​
Ground-based CMB experiments, such as the Atacama Cosmology Telescope (ACT), achieve high angular resolution thanks to their large telescope aperture, but they are limited in frequency coverage due to atmospheric emission and absorption. In contrast, space missions like Planck provide broad frequency coverage but have limited angular resolution because of the constrained telescope size that can be launched into space. In this paper, we proposed a multi-scale approach to component separation using needlet (spherical wavelet) decomposition, which enables the combination of multi-resolution data from heterogeneous CMB experiments to optimize the reconstruction of the thermal SZ effect from galaxy clusters. Using sky simulations, we demonstrated that combining Planck and ACT data through the wavelet-based component separation method NILC allows extraction of SZ emission from compact clusters that are invisible to Planck—thanks to ACT’s high-resolution channels—while simultaneously minimizing Galactic foreground contamination thanks to Planck’s high-frequency channels. This multi-resolution, multi-instrument approach effectively builds a "virtual ideal" experiment for component separation by exploiting the complementary strengths of each dataset. The method has since been applied to real data by the Planck and ACT collaborations, resulting in joint Planck-ACT thermal SZ maps and the discovery of new galaxy clusters (Aghanim et al., A&A 2019; Madhavacheril et al., PRD 2020; Coulton et al., PRD 2024).
​