Bayesian evidences for dark energy models in light of current observational data

We do a comprehensive study of the Bayesian evidences for a large number of dark energy models using a combination of latest cosmological data from SNIa, CMB, BAO, strong lensing time delay, growth measurements, measurements of Hubble parameter at different redshifts and measurements of angular diameter distance by Megamaser Cosmology Project. We consider a variety of scalar field models with different potentials as well as different parametrizations for the dark energy equation of state. Among 21 models that we consider in our study, we do not find strong evidences in favor of any evolving dark energy model compared to $\mathrm{\Lambda }\mathrm{CDM}$. For the evolving dark energy models, we show that purely nonphantom models have much better evidences compared to those models that allow both phantom and nonphantom behaviors. Canonical scalar field with exponential and tachyon field with square potential have highest evidences among all the models considered in this work. We also show that a combination of low redshift measurements decisively favors an accelerating $\mathrm{\Lambda }\mathrm{CDM}$ model compared to a nonaccelerating power law model.

Figure 1

$\mathrm{\Delta }\ln \mathcal{Z}$ for different models compared to $\mathrm{\Lambda }\mathrm{CDM}$ for four combinations of observational data. The horizontal line represents the $\mathrm{\Lambda }\mathrm{CDM}$ model. Models above this line have better evidences than $\mathrm{\Lambda }\mathrm{CDM}$ whereas models below this have lower evidences than $\mathrm{\Lambda }\mathrm{CDM}$. The blue, yellow, and green points represent models with insignificant, significant, and strong evidence compared to $\mathrm{\Lambda }\mathrm{CDM}$ according to Jeffrey’s scale given in Table 2.