Inferences for f(R) Models from Late-Time Megamaser Observational Data

Abstract

In this work, we study three widely used models of f(R) gravity, namely HuSawicki, Starobinsky and ArcTanh along with the standard cosmology model ($\Lambda$CDM). For this, we employ the megamaser angular diameter distance and velocity measurements from the Megamaser Cosmology Project, which provide a purely geometric determination of the Hubble constant. We constrain the parameters using the Markov Chain Monte Carlo method. Our results show that values of the Hubble Constant, $H_{0}$, obtained for all four models are in concordance with its value obtained from other late-time observational data such as SNe Ia. The constraints on $H_{0}$ in all the models under study are restrictive and the marginalized estimates lie close to 73 $\mathrm {km s^{-1} Mpc^{-1}}$. The marginalized estimates of the deviation parameter, b, for the three f(R) gravity models lie close to zero. This late-time dataset, thus predicts that f(R) models mimic $\Lambda$CDM. However, the matter density, $\Omega_m$, remains weakly constrained for all the models with its marginalized estimate close to 0.5. Further, comparison of the four models (f(R) models and $\Lambda$CDM) using information criteria such as Akaike Information Criterion and Bayesian Information Criterion shows that within current uncertainties, the dataset finds f(R) models statistically indistinguishable from $\Lambda$CDM. This is consistent with the fact that the favoured value of b for each of the f(R) models lies close to zero.