Is cosmic acceleration slowing down?

We investigate the course of cosmic expansion in its recent past using the Constitution SN Ia sample, along with baryon acoustic oscillations (BAO) and cosmic microwave background (CMB) data. Allowing the equation of state of dark energy (DE) to vary, we find that a coasting model of the universe ($q_0=0$) fits the data about as well as Lambda cold dark matter. This effect, which is most clearly seen using the recently introduced $Om$ diagnostic, corresponds to an increase of $Om$ and $q$ at redshifts $z\lesssim 0.3$. This suggests that cosmic acceleration may have already peaked and that we are currently witnessing its slowing down. The case for evolving DE strengthens if a subsample of the Constitution set consisting of $\mathrm{SNLS}+\mathrm{\text{ESSENCE}}+\mathrm{CfA}\mathrm{SN}\mathrm{Ia}$ data is analyzed in combination with $\mathrm{BAO}+\mathrm{CMB}$ data. The effect we observe could correspond to DE decaying into dark matter (or something else).

Figure 1

The $Om$ diagnostic is shown as a function of redshift for DE models with ${\Omega }_{0m}=0.27$ and $w=-1,-0.8,-0.6,-0.4,-0.2$ (bottom to top). For Phantom models (not shown) $Om$ would have the opposite curvature.

Figure 2

Reconstructed $q(z)$ and $Om(z)$ from $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}$ data (upper panels) and $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}+\mathrm{CMB}$ data (lower panels) using the CPL ansatz. Solid red lines show the best fit values of $Om(z)$ and $q(z)$ while dashed green lines show the $1\sigma$ C.L. Dotted blue lines represent the best fit spatially flat $\Lambda \mathrm{CDM}$ model. The dramatic difference between the upper panels and the lower ones is indicative of the inability of the CPL parametrization to fit the data at low and high redshifts simultaneously.

Figure 3

$1\sigma$ contours for CPL parameters $w_0\mathrm{\text{−}}{w}_1$ (left panel) and $w_0\mathrm{\text{−}}{\Omega }_{0m}$ (right panel) reconstructed using SN Ia data (red plusses), $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}$ data (green crosses) and $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}+\mathrm{CMB}$ data (blue stars). Light blue dot at $w_0=-1$, $w_1=0$ in the left panel represent spatially flat $\Lambda \mathrm{CDM}$ model. Note the consistency between SN Ia and BAO data. The absence of any overlap between the $1\sigma$ contours for $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}+\mathrm{CMB}$ and $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}$ data could be indicative of tension between the CPL parametrization and the data. We should note here that plotting small contours that indicate tight constraints on the parameters is misleading when the best fit result produces a bad fit to the data (blue contours).

Figure 4

The cosmological deceleration parameter $q(z)$ (left panel) and $Om(z)$ (right panel) reconstructed using a combination of SN Ia, BAO and CMB data and the ansatz (6). Solid red lines show best fit reconstructed results while dashed green lines show reconstructed results within $1\sigma$ C.L. Dotted blue lines show the best fit spatially flat $\Lambda \mathrm{CDM}$ model.

Figure 5

Top: $1\sigma$ contours for CPL parameters $w_0\mathrm{\text{−}}{w}_1$ (left panel) and $w_0\mathrm{\text{−}}{\Omega }_{0m}$ (right panel) reconstructed using SN Ia data $\mathrm{SNLS}+\mathrm{\text{ESSENCE}}+\mathrm{CfA}$, red plusses), $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}$ data (green crosses) and $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}+\mathrm{CMB}$ data (blue stars). Note the compatibility between the different data sets and that spatially flat $\Lambda \mathrm{CDM}$ (red cross at $w_0=-1$, $w_1=0$ in the left panel) appears to be in tension with this combination of data. Middle and bottom: Reconstructed $q(z)$ and $Om(z)$ from $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}$ data (middle). Reconstructed $q(z)$ and $Om(z)$ from $\mathrm{SN}\mathrm{Ia}+\mathrm{BAO}+\mathrm{CMB}$ data (bottom). In the middle and bottom panels solid red lines show the best fit values of $Om(z)$ and $q(z)$ while dashed green lines show the $1\sigma$ C.L. Dotted blue lines represent the best fit spatially flat $\Lambda \mathrm{CDM}$ model.