Model-independent determination of cosmic curvature based on Padé approximation

Abstract

Given observations of the standard candles and the cosmic chronometers, we apply Padé parameterization to the comoving distance and the Hubble paramter to find how stringent the constraint is set to the curvature parameter by the data. A weak informative prior is introduced in the modeling process to keep the inference away from the singularities. Bayesian evidence for different order of Padé parameterizations is evaluated during the inference to select the most suitable parameterization in light of the data. The data we used prefer a parameterization form of comoving distance as $D_{01}(z)=\frac{a_0 z}{1+b_1 z}$ as well as a competitive form $D_{02}(z)=\frac{a_0 z}{1+b_1 z + b_2 z^2}$. Similar constraints on the spatial curvature parameter are established by those models and given the Hubble constant as a byproduct: $Ω_k = 0.25^{+0.14}_{-0.13}$ (68\% confidence level [C.L.]), $H_0 = 67.7 \pm 2.0$ km/s/Mpc (68\% C.L.) for $D_{01}$, and $Ω_k = -0.01 \pm 0.13$ (68\% C.L.), $H_0 = 68.8 \pm 2.0$ km/s/Mpc (68\% C.L.) for $D_{02}$. The evidence of different models demonstrates the qualitative analysis of the Padé parameterizations for the comoving distance.