The 1% Concordance Hubble Constant

TL;DR

This work tests the consistency of Hubble constant determinations from CMB, BAO, and the cosmic distance ladder within flat ΛCDM. By combining WMAP9/ACT/SPT/BAO and distance-ladder priors and modeling the CMB likelihood with a Gaussian mixture, it derives a concordance value of $H_0 = 69.6 \pm 0.7$ ${\rm km\,s^{-1}\,Mpc^{-1}}$ and $\Omega_m = 0.286 \pm 0.008$, finding no strong evidence for new physics to reconcile the probes. The analysis shows all three data sources lie along a common degeneracy in the $H_0$–$D_V r_{d,\rm fid}/r_d$ plane and remains consistent across several data combinations. It also concludes that current data do not require extensions such as varying $N_{\rm eff}$ or $w \neq -1$, though future improvements in local $H_0$ precision and high-redshift BAO will sharpen tests of standard cosmology.

Abstract

The determination of the Hubble constant has been a central goal in observational astrophysics for nearly 100 years. Extraordinary progress has occurred in recent years on two fronts: the cosmic distance ladder measurements at low redshift and cosmic microwave background (CMB) measurements at high redshift. The CMB is used to predict the current expansion rate through a best-fit cosmological model. Complementary progress has been made with baryon acoustic oscillation (BAO) measurements at relatively low redshifts. While BAO data do not independently determine a Hubble constant, they are important for constraints on possible solutions and checks on cosmic consistency. A precise determination of the Hubble constant is of great value, but it is more important to compare the high and low redshift measurements to test our cosmological model. Significant tension would suggest either uncertainties not accounted for in the experimental estimates, or the discovery of new physics beyond the standard model of cosmology. In this paper we examine in detail the tension between the CMB, BAO, and cosmic distance ladder data sets. We find that these measurements are consistent within reasonable statistical expectations, and we combine them to determine a best-fit Hubble constant of 69.6+/-0.7 km/s/Mpc. This value is based upon WMAP9+SPT+ACT+6dFGS+BOSS/DR11+H_0/Riess; we explore alternate data combinations in the text. The combined data constrain the Hubble constant to 1%, with no compelling evidence for new physics.

Figures (7)

• Figure 1: WMAP9-based (top, middle) and Planck-based (bottom) CMB data sets for a flat $\Lambda$CDM cosmological model, compared to local $H_0$ measurements and BAO data at $z=0.57$. Gray points in the top panel are from this paper's WMAP9/2014 chain; those in the middle are from (WMAP9+ACT+SPT)/2014. Gray points in the bottom panel are from the Planck chains planck/16:2013 which include the (ACT+SPT)/2013 data combination (see text and Table 2). The peak of each 2D-marginalized CMB likelihood is indicated with a black filled circle and the contours for the chain points enclose 68% and 95% of the probability, as computed from a 10-component 2D Gaussian mixture fit (see text). The BAO and $H_0$ measurements are indicated by vertical and horizontal lines, respectively, with dashed lines showing $1\sigma$ error bars. The gray ellipses show 68% and 95% contours for the product of the BAO and $H_0$ likelihoods. The peak of the combined CMB+BAO+$H_0$ likelihoods is indicated with an open square, and its value is noted in each panel.
• Figure 2: Combination of $z=0.57$ BAO, local $H_0$, and CMB data. Gray points and contours are from (WMAP9+ACT+SPT)/2014, as described in the text. The black filled circle is the peak of the 2D marginalized CMB likelihood. The concordance value, the open square, is maximum likelihood of the product of the 2D CMB, BAO, and $H_0$ likelihoods.
• Figure 2: (Continued)
• Figure 3: Combination of $z=0.57$ BAO, $H_0$, and CMB data, as in Figure \ref{['fig:bao_h0_cmb1']}, but now with additional cosmological parameters not required by the data such as space curvature and/or an equation of state parameter $w \ne -1$, and/or massive neutrinos. For models only weakly constrained by CMB data the 2D Gaussian mixture approach used is not ideal so the contours should be viewed with caution.
• Figure 3: (Continued)