The Atacama Cosmology Telescope: A Measurement of the Cosmic Microwave Background Power Spectra at 98 and 150 GHz

TL;DR

This paper reports Atacama Cosmology Telescope DR4 measurements of CMB temperature and polarization power spectra at 98 and 150 GHz from 2013–2016 nighttime surveys. It introduces a rigorous blinding workflow, extensive null and consistency tests, and a curved-sky power-spectrum pipeline with a comprehensive covariance model that includes lensing, super-sample variance, and foregrounds. The ACTPol+DR4 analysis yields ACT-only $\Lambda$CDM parameter constraints (notably $H_0\approx67.6$ km s$^{-1}$ Mpc$^{-1}$) with strong support from TE/EE data, while cross-checks with Planck show general agreement though some TE discrepancies remain. The work provides a publicly released, foreground-aware CMB spectrum and foreground parameter posteriors, and demonstrates robust consistency with standard cosmology while illustrating the value of multi-frequency, polarization-dominated measurements for precision cosmology.

Abstract

We present the temperature and polarization angular power spectra of the CMB measured by the Atacama Cosmology Telescope (ACT) from 5400 deg$^2$ of the 2013-2016 survey, which covers $>$15000 deg$^2$ at 98 and 150 GHz. For this analysis we adopt a blinding strategy to help avoid confirmation bias and, related to this, show numerous checks for systematic error done before unblinding. Using the likelihood for the cosmological analysis we constrain secondary sources of anisotropy and foreground emission, and derive a "CMB-only" spectrum that extends to $\ell=4000$. At large angular scales, foreground emission at 150 GHz is $\sim$1% of TT and EE within our selected regions and consistent with that found by Planck. Using the same likelihood, we obtain the cosmological parameters for $Λ$CDM for the ACT data alone with a prior on the optical depth of $τ=0.065\pm0.015$. $Λ$CDM is a good fit. The best-fit model has a reduced $χ^2$ of 1.07 (PTE=0.07) with $H_0=67.9\pm1.5$ km/s/Mpc. We show that the lensing BB signal is consistent with $Λ$CDM and limit the celestial EB polarization angle to $ψ_P =-0.07^{\circ}\pm0.09^{\circ}$. We directly cross correlate ACT with Planck and observe generally good agreement but with some discrepancies in TE. All data on which this analysis is based will be publicly released.

Figures (27)

• Figure 1: The average $98\,$ GHz (cyan) and $150\,$ GHz (black) beam profiles in "gain above isotropic" ($4\pi/\Omega_B$). The forward gains are 74.5 and 78.4 dBi respectively. The two dashed curves on the bottom show the scattering beam due to the surface roughness. For reference, the blue dash-dot line, offset for clarity, shows the slope of a $1/\theta^3$ profile. Negative values due to noise fluctuations are not plotted.
• Figure 2: The cumulative ACT DR4 coverage shown in equatorial coordinates for observations between 2013--2016. The background is the Planck 353 GHz intensity map. The x-axis (y-axis) shows the RA (dec.) coordinates in degrees. The color scale indicates the depth of the map. The noise levels are reported in Table \ref{['tab:obs']}. Regions w2, w6, and D8 are not part of the cosmological analysis.
• Figure 3: Graphical depiction of obtaining the spatial window function for the D6 region. The top panel shows the cross-linking map; the middle panel shows the normalized inverse noise variance of the coverage; and the bottom panel shows the spatial window function. The color scales are from 0 to 1 shown in blue to red. After applying the cross-linking threshold to the top map and the noise threshold to the middle map, they are multiplied together to obtain the bottom map. The source mask and Planck Galactic mask, not shown, are also applied. See Figure \ref{['fig:actpol_coverage']} for the size and location of the region. The outline of the bottom map represents the "Area" in Table \ref{['tab:obs']}, whereas "Area PS" corresponds to the effective area after inverse noise variance weighting.
• Figure 4: A comparison of the modeled Fourier-Space filter transfer function to simulations. Using signal-only CMB maps of the D56 region, we compute the power spectra before and after applying the Fourier space filter (cutting $|\ell_x| < 90$ and $|\ell_y| < 50$). The binned transfer function for TT, and the EE/BB transfer matrix elements at each $\ell$ bin are computed using Equation \ref{['eq:tfunc']}, shown in all three panels. Note that the bottom two panels show an expanded view of the $\ell<1000$ region in the top panel. The analytic estimate of the transfer function is shown in gray solid line. The model describes the simulations to within 0.5% for $\ell>300$. The bottom panel shows the EE/BB mixing term, which is $<0.15$% for $\ell>350$. This mixing correction was necessary only for the sensitivity levels achieved in D56.
• Figure 5: The EB null angle $\psi_P$ is shown for all fields/seasons/arrays of data. The weighted mean of 150 GHz (98 GHz) angles is $-0.07^{\circ}\pm0.09^{\circ}$ ($-0.11^{\circ}\pm0.15^{\circ}$), and $\chi^2/\mathrm{dof} = 1.20$ (0.68).