Tale of stable interacting dark energy, observational signatures, and the $H_0$ tension

TL;DR

This paper explores stable interacting dark energy models with observational data, finding a mild preference for phantom dark energy and partial alleviation of the H0 tension, but overall favoring the standard Lambda-CDM model.

Contribution

It introduces two novel stable IDE models derived from known unstable ones and provides observational constraints, addressing the H0 tension within these models.

Findings

Preference for phantom dark energy with w_x < -1

Partial alleviation of the H0 tension

Standard Lambda-CDM model remains preferred

Abstract

We investigate the observational consequences of a novel class of stable interacting dark energy (IDE) models, featuring interactions between dark matter (DM) and dark energy (DE). In the first part of our work, we start by considering two IDE models which are known to present early-time linear perturbation instabilities. Applying a transformation depending on the dark energy equation of state (EoS) to the DM-DE coupling, we then obtain two novel stable IDE models. Subsequently, we derive robust and accurate constraints on the parameters of these models, assuming a constant EoS $w_x$ for the DE fluid, in light of some of the most recent publicly available cosmological data. These include Cosmic Microwave Background (CMB) temperature and polarization anisotropy measurements from the \textit{Planck} satellite, a selection of Baryon Acoustic Oscillation measurements, Supernovae Type-Ia luminosity distance measurements from the JLA sample, and measurements of the Hubble parameter up to redshift $2$ from cosmic chronometers. Our analysis displays a mild preference for the DE fluid residing in the phantom region ($w_x<-1$), with significance up to 95\% confidence level, while we obtain new upper limits on the coupling parameter between the dark components. The preference for a phantom DE suggests a coupling function $Q<0$, thus a scenario where energy flows from the DE to the DM. We also examine the possibility of addressing the $H_0$ and $\sigma_8$ tensions, finding that only the former can be partially alleviated. Finally, we perform a Bayesian model comparison analysis to quantify the possible preference for the two IDE models against the standard concordance $\Lambda$CDM model, finding that the latter is always preferred with the strength of the evidence ranging from positive to very strong.