Observational constraints on an interacting dark energy model

TL;DR

This study investigates a simple interacting dark energy model in a flat cosmology, where dark matter decays into dark energy at a constant rate $\Gamma$ and the momentum transfer is fixed covariantly to ensure consistent perturbations. A time-varying dark energy equation of state, $w_{de}=w_0 a+w_e(1-a)$, is crucial to avoid early-time instabilities and yields viable cosmologies when combined with data. Using Monte Carlo Markov Chain analyses with WMAP5, SN Union, and BAO data, the authors show a strong degeneracy between $\Gamma$ and present densities in the CMB, but find that SN and BAO data constrain the interaction to roughly $-0.23<Γ/H_0<+0.15$ (95% CL); the dominant observable remains the ISW effect, motivating future CMB–LSS cross-correlation studies. The work documents the need for momentum-transfer specification, provides a modified CAMB/CosmoMC toolkit for the model, and outlines how ISW measurements could tighten constraints beyond what Planck-like CMB data achieve alone.

Abstract

We use observations of cosmic microwave background anisotropies, supernova luminosities and the baryon acoustic oscillation signal in the galaxy distribution to constrain the cosmological parameters in a simple interacting dark energy model with a time-varying equation of state. Using a Monte Carlo Markov Chain technique we determine the posterior likelihoods. Constraints from the individual data sets are weak, but the combination of the three data sets confines the interaction constant $Γ$ to be less than 23% of the expansion rate of the Universe $H_0$; at 95% CL $-0.23 < Γ/H_0 < +0.15$. The CMB acoustic peaks can be well fitted even if the interaction rate is much larger, but this requires a larger or smaller (depending on the sign of interaction) matter density today than in the non-interacting model. Due to this degeneracy between the matter density and the interaction rate, the only observable effect on the CMB is a larger or smaller integrated Sachs-Wolfe (ISW) effect. While SN or BAO data alone do not set any direct constraints on the interaction, they exclude the models with very large matter density, and hence indirectly constrain the interaction rate when jointly analysed with the CMB data. To enable the analysis described in this paper, we present in a companion paper [arXiv:0907.4981] a new systematic analysis of the early radiation era solution to find the adiabatic initial conditions for the Boltzmann integration.

Figures (10)

• Figure 1: Marginalized likelihoods for the interacting model with WMAP, WMAP&ACBAR, and WMAP&SN&BAO data. The curves show 68% and 95% CL regions: dotted magenta/grey for WMAP, solid cyan/grey for WMAP&ACBAR, and solid black for WMAP&SN&BAO data.
• Figure 2: Models with most of their parameters equal to the parameters of the best-fitting to WMAP non-interacting model.
• Figure 3: Redshift evolution of the ISW source. Dashed vertical lines indicate last scattering ($z \approx 1090$) and the time when the interaction starts to directly modify the evolution of the ISW source ($z \approx 10$).
• Figure 4: 2d marginalized likelihoods for the interacting model with only the WMAP, only the SN, and only the BAO data. The darker blue or red colours indicate 68% CL regions while the lighter blue or red colours indicate 95% CL regions with the SN or BAO data, respectively. The best-fits stand for the best-fitting models in the ranges shown in this figure. Therefore, here the BAO best-fitting model differs from the tabulated one which is at $\Gamma/H_0 = 2.92$; see Table \ref{['table:Models']}.
• Figure 5: 2d marginalized likelihoods for the interacting model with the WMAP, WMAP&SN, WMAP&BAO and WMAP&SN&BAO data.