[Paper Review] A grounded perspective on New Early Dark Energy using ACT, SPT, and BICEP/Keck

Juan S. Cruz, Florian Niedermann+1 more|arXiv (Cornell University)|Sep 6, 2022

Cosmology and Gravitation Theories4 citations

TL;DR

This paper evaluates the New Early Dark Energy (NEDE) model using updated CMB data from ACT, SPT, and BICEP/Keck, along with the latest S$H_{0}$ES local Hubble constant prior. It finds strong statistical support for NEDE, reducing the Hubble tension by improving the fit over ΛCDM by Δχ² = -15.9 (ACT) and -23.1 (SPT), with a 4.8σ preference for non-zero NEDE fraction and a best-fit H₀ of 72.09 km/s/Mpc.

ABSTRACT

We examine further the ability of the New Early Dark Energy model (NEDE) to resolve the current tension between the Cosmic Microwave Background (CMB) and local measurements of $H_0$ and the consequences for inflation. We perform new Bayesian analyses, including the current datasets from the ground-based CMB telescopes Atacama Cosmology Telescope (ACT), the South Pole Telescope (SPT), and the BICEP/Keck telescopes, employing an updated likelihood for the local measurements coming from the S$H_0$ES collaboration. Using the S$H_0$ES prior on $H_0$, the combined analysis with Baryonic Acoustic Oscillations (BAO), Pantheon, Planck and ACT improves the best-fit by $Δχ^2 = -15.9$ with respect to $Λ$CDM, favors a non-zero fractional contribution of NEDE, $f_{ m NEDE} > 0$, by $4.8σ$, and gives a best-fit value for the Hubble constant of $H_0 = 72.09$ km/s/Mpc (mean $71.48_{-0.81}^{+0.79}$ with $68\%$ C.L.). A similar analysis using SPT instead of ACT yields consistent results with a $Δχ^2 = - 23.1$ over $Λ$CDM, a preference for non-zero $f_{ m NEDE}$ of $4.7σ$ and a best-fit value of $H_0=71.77$ km/s/Mpc (mean $71.43_{-0.84}^{+0.84}$ with $68\%$ C.L.). We also provide the constraints on the inflation parameters $r$ and $n_s$ coming from NEDE, including the BICEP/Keck 2018 data, and show that the allowed upper value on the tensor-scalar ratio is consistent with the $Λ$CDM bound, but, as also originally found, with a more blue scalar spectrum implying that the simplest curvaton model is now favored over the Starobinsky inflation model.

Motivation & Objective

• To assess whether the New Early Dark Energy (NEDE) model can resolve the Hubble tension between CMB and local measurements of H₀.
• To evaluate the consistency of NEDE with recent high-precision CMB datasets from ACT, SPT, and BICEP/Keck.
• To examine the implications of NEDE for inflation parameters, particularly the tensor-to-scalar ratio r and scalar spectral index nₛ.
• To test whether the NEDE model improves the global fit to cosmological data compared to ΛCDM, including Planck, BAO, Pantheon, and local H₀ measurements.

Proposed method

• Performing Bayesian MCMC analyses using updated likelihoods for ACT, SPT, and BICEP/Keck 2018 CMB data.
• Incorporating the latest S$H_{0}$ES 2021 prior on H₀ (73.04 ± 1.04 km/s/Mpc) as a constraint in the analysis.
• Comparing the NEDE model against the standard ΛCDM model using Δχ² statistics to assess statistical preference.
• Including Baryonic Acoustic Oscillations (BAO) and Pantheon supernova data to constrain the Hubble constant and sound horizon.
• Analyzing the impact of NEDE on inflation parameters r and nₛ, particularly using BICEP/Keck 2018 data on tensor modes.
• Using the Q_dmap statistic to assess the global goodness-of-fit and consistency of the data with the models.

Experimental results

Research questions

• RQ1 Does the NEDE model significantly improve the fit to cosmological data compared to ΛCDM when including ACT, SPT, and BICEP/Keck CMB data?
• RQ2 To what extent does NEDE alleviate the Hubble tension, as quantified by the H₀ discrepancy between CMB and local measurements?
• RQ3 What constraints does the NEDE model place on the tensor-to-scalar ratio r and scalar spectral index nₛ, and how do they compare to ΛCDM expectations?
• RQ4 Is the preference for NEDE robust across different CMB datasets, such as ACT versus SPT?
• RQ5 How does the inclusion of the S$H_{0}$ES 2021 H₀ prior affect the statistical significance of the NEDE model?

Key findings

• The NEDE model improves the global fit to data by Δχ² = -15.9 over ΛCDM when combining Planck, BAO, Pantheon, and ACT data.
• With the S$H_{0}$ES 2021 prior on H₀, the NEDE model shows a 4.8σ preference for a non-zero fractional contribution f_NEDE > 0.
• The best-fit Hubble constant under NEDE is H₀ = 72.09 km/s/Mpc, with a 68% credible interval of 71.49 ± 0.82 km/s/Mpc.
• Using SPT instead of ACT yields a Δχ² = -23.1 over ΛCDM, with a 4.7σ preference for f_NEDE > 0 and a best-fit H₀ = 71.77 km/s/Mpc (71.43 ± 0.85 km/s/Mpc at 68% C.L.).
• The NEDE model leads to a more blue-tilted scalar spectrum (higher nₛ), favoring the simplest curvaton model over the Starobinsky inflation model.
• The upper bound on the tensor-to-scalar ratio r remains consistent with the ΛCDM limit, but the improved fit to data supports a slightly higher nₛ in the NEDE framework.