Elucidating cosmological model dependence with H_0

Eoin Ó Colgáin,M M Sheikh-Jabbari

doi:10.1140/epjc/s10052-021-09708-2

Abstract

We observe that the errors on the Hubble constant H_0, a universal parameter in any FLRW cosmology, can be larger in specific cosmological models than Gaussian processes (GP) data reconstruction. We comment on the prior mean function and trace the smaller GP errors to stronger correlations, which we show precludes all well studied dynamical dark energy models. We also briefly illustrate cosmographic expansions as another model independent cosmological reconstruction. Our analysis suggests that “cosmological model independence”, especially in the statement of Hubble tension, has become a misnomer.

Highlights

We show that while the correlations in simpler models such as flat ΛCDM and wCDM are typically stronger than the Gaussian Processes (GP) output, in turn correlations from GP are generically stronger than dynamical dark energy (DDE) models
The errors on H0 for both ΛCDM and wCDM are within the GP errors, but for CPL we find that the error on H0 is larger
We started with some observations on Taylor expansion and the regime where the corresponding cosmographic expansion may be regarded as model independent in a bona fide sense

Summary

Introduction

Cosmological tensions [1] point to problems with the assumption that the Universe is flat ΛCDM. Some assumptions underlying supernovae are in a state of flux [2,3,4], and even the assumption that the Universe is isotropic and homogeneous is being called into question [5,6,7,8,9,10] This perpetual cycle of assumptions and contradictions is integral to cosmology. Since the two determinations of H0 are independent of the cosmological model, it is an immediate corollary that Hubble tension has no cosmological resolution, at least within the Friedmann–Lemaître–Robertson–Walker (FLRW) framework This may be true, so there may be no contradiction. GP represents a restriction on the parameter space of DDE models

Taylor expansion

Gaussian processes

Mean function

Analysis

Correlations