Observational tests of non-adiabatic Chaplygin gas

TL;DR

This paper tests a non-adiabatic Chaplygin gas model against cosmological observations, finding a best-fit parameter indicating energy transfer from dark energy to dark matter, which challenges the standard Lambda-CDM model.

Contribution

It provides observational constraints on the non-adiabatic Chaplygin gas model, demonstrating its viability and differences from the standard cosmological model using multiple data sets.

Findings

Best-fit alpha around -0.5 indicating energy transfer from dark energy to dark matter

Lambda-CDM model (alpha=0) is outside 3-sigma confidence interval

Model fits supernovae, CMB, and large-scale structure data consistently

Abstract

In a previous paper it was shown that any dark sector model can be mapped into a non-adiabatic fluid formed by two interacting components, one with zero pressure and the other with equation-of-state parameter $\omega = -1$. It was also shown that the latter does not cluster and, hence, the former is identified as the observed clustering matter. This guarantees that the dark matter power spectrum does not suffer from oscillations or instabilities. It applies in particular to the generalised Chaplygin gas, which was shown to be equivalent to interacting models at both background and perturbation levels. In the present paper we test the non-adiabatic Chaplygin gas against the Hubble diagram of type Ia supernovae, the position of the first acoustic peak in the anisotropy spectrum of the cosmic microwave background and the linear power spectrum of large scale structures. We consider two different samples of SNe Ia, namely the Constitution and SDSS compilations, both calibrated with the MLCS2k2 fitter, and for the power spectrum we use the 2dFGRS catalogue. The model parameters to be adjusted are the present Hubble parameter, the present matter density and the Chaplygin gas parameter $\alpha$. The joint analysis best fit gives $\alpha \approx - 0.5$, which corresponds to a constant-rate energy flux from dark energy to dark matter, with the dark energy density decaying linearly with the Hubble parameter. The $\Lambda$CDM model, equivalent to $\alpha = 0$, stands outside the $3\sigma$ confidence interval. This result is still valid if we use, as supernovae samples, the SDSS and Union2.1 compilations calibrated with the SALT2 fitter.