Inferences for f(R) Models from Late-Time Megamaser Observational Data
📄 arXiv:2607.21129 · 📥 PDF · 2026-07-23 · astro-ph.CO
Authors: Umang [arXiv · scholar] , Abha Dev Habib [arXiv · scholar] , Nisha Rani [arXiv · scholar]
🕰 Orloj analysis
Tato práce zkoumá tři modely f(R) gravitace (Hu-Sawicki, Starobinsky, ArcTanh) a standardní model ΛCDM pomocí dat z megamaserů pro určení Hubbleovy konstanty. Zjišťuje, že modely f(R) jsou statisticky nerozlišitelné od ΛCDM a jejich parametry odchylky se blíží nule, přičemž H0 je v souladu s jinými pozdními daty.
💡 Práce představuje solidní aplikaci existujících modelů na novější datový soubor, ale její závěry o podobnosti f(R) modelů s ΛCDM jsou konzistentní s mnoha předchozími studiemi, což snižuje její celkovou novost a přidanou hodnotu pro rozlišení teorií.
Categories:
THE-18
GRV-4
THE-1
COS-1
MET-1
EXP-2
✓ falsifiable, limit_reductions, modest_claims
⚠ matter_density_weakly_constrained, code_not_available, no_pre_registration
📄 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 ($Λ$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 $Λ$CDM. However, the matter density, $Ω_m$, 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 $Λ$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 $Λ$CDM. This is consistent with the fact that the favoured value of b for each of the f(R) models lies close to zero.