Evidence of galaxy cluster rotation in the cosmic microwave background

TL;DR

This work presents the first robust detection of the rotational kinematic Sunyaev–Zel’dovich (rkSZ) effect in nearby galaxy clusters by orienting Planck CMB temperature maps using rotation directions inferred from cluster member galaxies. The authors implement an improved analysis pipeline that combines tSZ-deprojected NILC maps with an E-mode subtraction to suppress primary CMB fluctuations, and perform an oriented, weighted stacking of 25 X-ray–matched clusters. They report a dipole signature aligned with the rotation direction at $3.6\sigma$, with a mean amplitude around $25\ \mu$K and a best-fit rkSZ amplitude $A_{\rm rkSZ} = 1.05 \pm 0.32$, broadly consistent with solid-body rotation models and hydrodynamical simulations. This result introduces rkSZ as a new, practical tool to probe the dynamical state and rotational support of the intracluster medium, with implications for understanding hydrostatic mass bias and cluster evolution in the low-redshift universe. $\Delta T_{\rm rkSZ}$ observations, combined with future higher-resolution CMB data, will enable more precise constraints on cluster rotation across masses and dynamical states.

Abstract

We report the first robust evidence for the rotational kinematic Sunyaev-Zel'dovich (rkSZ) effect, produced by the Thomson scattering of cosmic microwave background (CMB) photons off rotating intracluster gas. By combining CMB intensity and polarization measurements from the $\it{Planck}$ satellite with spectroscopic member-galaxy redshifts from the Sloan Digital Sky Survey in a sample of 25 X-ray cross-matched, low-redshift ($0.02< z< 0.09)$, massive ($10^{13.9}\lesssim M_{\rm 500c}/M_\odot \lesssim 10^{14.6}$) galaxy clusters, we detect a dipolar rkSZ signature aligned with the estimated rotation direction of each cluster, ruling out a chance fluctuation at 99.98% confidence (3.6$σ$). The significance of this measurement is enhanced by several new methodological improvements for isolating the rkSZ signal from primary CMB fluctuations and noise. The amplitude and shape of the signal are qualitatively consistent with predictions from state-of-the-art hydrodynamical simulations. These results establish a new tool with which to probe the dynamical state of galaxy clusters using CMB data.

Figures (15)

• Figure 1: Top: oriented and weighted stack of filtered and cleaned Planck CMB temperature data on 25 galaxy clusters. The cutouts are rescaled in units of $R_{\rm 500c}$ and aligned with the estimated cluster rotation direction, with the right side ($x \geq 0$) rotating towards the observer, as indicated by the white arrow. A dipole signal is clearly visible along the expected rotation direction. Bottom: significance of the dipole estimated from the oriented stack (red line) compared with the noise distribution obtained by repeating the weighted stack 30,000 times at random locations. The measured dipole amplitude exceeds 99.98% of the random realizations, indicating strong evidence ($3.6\sigma$) of cluster rotation. The black dashed line shows the best-fit Gaussian to the noise.
• Figure 2: Azimuthally-averaged rkSZ profile derived from our weighted stack of 25 clusters. The red curve shows a theoretical prediction based on a solid-body rotation model and the Battaglia Battaglia:2016xbi electron number density profile. We find significant evidence for a non-zero amplitude, $A_{\rm rkSZ}=1.05\pm 0.32$. Note that the data points are highly correlated.
• Figure 3: Masks and main CMB maps used in this work. The top-left panel shows the preprocessing mask used for inpainting the multi-frequency maps prior to component separation. It is constructed from the union of a Galactic plane mask and the Planck LFI and HFI point-source masks. The top right panel shows the analysis mask, defined as the union of the inpainting mask and the Planck polarization confidence mask, and subsequently apodized with a 30-arcminute NaMaster C2 apodization. The red shaded region indicates the subset of the SDSS footprint from which our cluster catalog is drawn. We use the union of this footprint and the analysis mask when performing random stacks when determining the significance of our measurements. The bottom-left shows the full-sky tSZ-deprojected blackbody temperature NILC map constructed from multi-frequency Planck intensity measurements with pyilc. The bottom-right panel shows the E-mode-cleaned tSZ-deprojected blackbody temperature NILC map, $\tilde{T}$, which we use for our fiducial analysis. As discussed in the text, the $\tilde{T}$ map is constructed from maps that have already been masked to ensure numerical stability in the spherical-harmonic transforms.
• Figure 4: Left: properties of the 13 cosine needlets used in our NILC maps, showing each filter's peak multipole, real-space smoothing scale, and excluded Planck channels. Right: harmonic-space needlet filters $h^{(I)}_\ell$ used in this work.
• Figure 5: Left: ratio of the power spectrum of the E-mode-subtracted NILC blackbody temperature map, $\tilde{T}$, to that of the raw NILC $T$ map. A cosmic-variance-limited E-mode survey could achieve significantly better variance reduction (black curve). Right: harmonic-space filter applied to remove large- and small-scale modes prior to oriented stacking.