Skip to main content
Publication

Constraints on CDM Extensions from the SPT-3G 2018 EE and TE Power Spectra

Authors

Balkenhol, L. ; Dutcher, D.; Ade, P. A. R. ; Ahmed, Z. ; Anderes, E. ; Anderson, A. J. ; Archipley, M. ; Avva, J. S. ; Aylor, K. ; Barry, P.; Bender, A.; Bleem, L.; Byrum, K.; Cecil, T.; Chang, C.; Carter, F.; Gualtieri, R.; Henning, J.; Kuhlmann, S.; Pan, Z.; Wang, G.; Yefremenko, V.

Abstract

We present constraints on extensions to the Lambda CDM cosmological model from measurements of the E-mode polarization autopower spectrum and the temperature-E-mode cross-power spectrum of the cosmic microwave background (CMB) made using 2018 SPT-3G data. The extensions considered vary the primordial helium abundance, the effective number of relativistic degrees of freedom, the sum of neutrino masses, the relativistic energy density and mass of a sterile neutrino, and the mean spatial curvature. We do not find clear evidence for any of these extensions, from either the SPT-3G 2018 dataset alone or in combination with baryon acoustic oscillation and Planck data. None of these model extensions significantly relax the tension between Hubble-constant, H-0, constraints from the CMB and from distance-ladder measurements using Cepheids and supernovae. The addition of the SPT-3G 2018 data to Planck reduces the square-root of the determinants of the parameter covariance matrices by factors of 1.3-2.0 across these models, signaling a substantial reduction in the allowed parameter volume. We also explore CMB-based constraints on H-0 from combined SPT, Planck, and ACT DR4 datasets. While individual experiments see some indications of different H-0 values between the TT, TE, and EE spectra, the combined H-0 constraints arc consistent between the three spectra. For the full combined datasets, we report H-0 = 67.49 +/- 0.53 km s(-1) Mpc(-1), which is the tightest constraint on H-0 from CMB power spectra to date and in 4.1 sigma tension with the most precise distance-ladder-based measurement of H-0. The SPT-3G survey is planned to continue through at least 2023, with existing maps of combined 2019 and 2020 data already having similar to 3.5 x lower noise than the maps used in this analysis.