Constraints on coupled dark energy using CMB data from WMAP and SPT

TL;DR

The paper investigates a dark-sector coupling between a dynamical dark-energy scalar field and dark matter, implemented as a constant coupling $β$ in a scalar-tensor-like framework. Using CMB observations from WMAP7 and SPT, it constrains $β$ through a modified CAMB/IDEA implementation integrated with COSMOMC, accounting for CMB-lensing, curvature, and massive neutrinos. The results show $β<0.063$ (68% CL) and $<0.11$ (95% CL) with WMAP7+SPT, with modest shifts when extending the parameter space (e.g., $N_{eff}$, $A_L$, $Ω_K$, $f_ u$); a small peak near $β=0.041$ emerges when including external priors (HST/BAO/SN Ia) but remains compatible with zero at 1σ. Forecasts using Planck+SPT mock data indicate that future observations could constrain $β$ to better than $1 ext{%}$ and determine whether the current marginal nonzero coupling is genuine.

Abstract

We consider the case of a coupling in the dark cosmological sector, where a dark energy scalar field modifies the gravitational attraction between dark matter particles. We find that the strength of the coupling β is constrained using current Cosmic Microwave Background (CMB) data, including WMAP7 and SPT, to be less than 0.063 (0.11) at 68% (95%) confidence level. Further, we consider the additional effect of the CMB-lensing amplitude, curvature, effective number of relativistic species and massive neutrinos and show that the bound from current data on β is already strong enough to be rather stable with respect to any of these variables. The strongest effect is obtained when we allow for massive neutrinos, in which case the bound becomes slightly weaker, β < 0.084(0.14). A larger value of the effective number of relativistic degrees of freedom favors larger couplings between dark matter and dark energy as well as values of the spectral index closer to 1. Adding the present constraints on the Hubble constant, as well as from baryon acoustic oscillations and supernovae Ia, we find β < 0.050(0.074). In this case we also find an interesting likelihood peak for β = 0.041 (still compatible with 0 at 1σ). This peak comes mostly from a slight difference between the Hubble parameter HST result and the WMAP7+SPT best fit. Finally, we show that forecasts of Planck+SPT mock data can pin down the coupling to a precision of better than 1% and detect whether the marginal peak we find at small non zero coupling is a real effect.

Figures (17)

• Figure 1: CMB TT temperature spectra for three values of $\beta$. Data are taken from WMAP7 Larson:2010gs.
• Figure 2: CMB TT temperature spectra (top panel) and dimensionless lensing potential $l(l+1)C^{\phi \phi}/(2\pi)$ versus multipole $l$ for three values of the coupling $\beta$ .
• Figure 3: WMAP7 and SPT data used for this work. SZ effects are not included in the best fit line (solid black) (see fig.5 of k11 for the best fit including SZ). The best fit from WMAP+SPT, for a $\Lambda$CDM without foregrounds is also shown (dash-dotted black line). This plot is similar to fig.5 of k11 (without SZ in the best fit); we reproduced it here for convenience and because it will allow us to neglect SZ when combining Planck with mock SPT data, in section V. All foregrounds are instead included when SPT is combined with WMAP7.
• Figure 4: WMAP7 and SPT data used for this work, with an extra factor of $\ell^2$. SZ effects are not included in the best fit line (solid black) (see fig.5 of k11 for the best fit including SZ). The best fit from WMAP+SPT, for a $\Lambda$CDM without foregrounds is also shown (dash-dotted black line).
• Figure 5: Confidence contours for the cosmological parameters for coupled quintessence models. We compare runs cq1 (red) and cq2 (white). The light blue asterisks mark the best fit points for $cq1$. 1-sigma and 2-sigma contours are shown.