New analysis of SNeIa Pantheon Catalog: Variable speed of light as an alternative to dark energy

TL;DR

This paper proposes that a variable speed of light in an Einstein-de Sitter universe can explain late-time cosmic acceleration observed in supernova data, eliminating the need for dark energy and aligning with previous CMB analyses.

Contribution

It introduces a novel variable speed of light model that fits supernova data better than Lambda-CDM without dark energy, reconciling early and late universe observations.

Findings

Achieves a superior fit to SNeIa Pantheon data without dark energy.

Finds a consistent Hubble constant around 47 km/s/Mpc.

Explains cosmic acceleration through declining speed of light.

Abstract

In A&A 412, 35 (2003) Blanchard, Douspis, Rowan-Robinson, and Sarkar (BDRS) slightly modified the primordial fluctuation spectrum and produced an excellent fit to WMAP's CMB power spectrum for an Einstein-de Sitter (EdS) universe, bypassing dark energy. Curiously, they obtained a Hubble value of $H_0\approx46$, in sharp conflict with the canonical range $H_0\sim67-73$. However, we will demonstrate that the reduced value of $H_0\approx46$ achieved by BDRS is fully compatible with the use of variable speed of light in analyzing the late-time cosmic acceleration observed in Type Ia supernovae (SNeIa). In Phys. Lett. B 862, 139357 (2025) we uncovered a hidden aspect in a generic class of scale-invariant actions: the dynamics of the dilaton can induce a variation in the speed of light as $c\propto\chi^{1/2}$, causing $c$ to vary alongside $\chi$ across spacetime. For an EdS universe with varying $c$, besides the effects of cosmic expansion, light waves emitted from distant SNeIa are further subject to a refraction effect, which alters the Lemaitre redshift relation to $1+z=a^{-3/2}$. Based on this new formula, we achieve a fit to the SNeIa Pantheon Catalog exceeding the quality of the $\Lambda$CDM model. Crucially, our approach does not require dark energy and produces $H_0=47.2$ in strong alignment with the BDRS finding of $H_0\approx46$. Hence, BDRS's analysis of the (early-time) CMB power spectrum and our variable-$c$ analysis of the (late-time) Hubble diagram of SNeIa fully agree on two counts: (i) the dark energy hypothesis is avoided, and (ii) $H_0$ is reduced to $\sim47$, which also yields an age $t_0=2/(3H_0)=13.8$ Gy for an EdS universe, without requiring dark energy. Most importantly, we will demonstrate that the late-time acceleration can be attributed to the declining speed of light in an expanding EdS universe, rather than to a dark energy component.