Cosmic chronometers, Pantheon+ supernovae, and quasars favor coasting cosmologies over the flat $\Lambda$CDM model

TL;DR

This study compares coasting cosmological models with curvature to the flat $\\Lambda$CDM model using diverse observational data, finding that coasting models are generally favored and highlighting issues with overfitting and data scatter.

Contribution

It provides a comprehensive comparison of coasting and flat $\\Lambda$CDM models using multiple datasets and introduces new insights into model fitting and data consistency.

Findings

Coasting models are generally favored over flat $\\Lambda$CDM.

Overfitting issues are identified in the flat $\\Lambda$CDM model with Pantheon+ SNe.

Large intrinsic scatter in QSO data suggests the need for refined error estimates.

Abstract

We test and compare coasting cosmological models with curvature parameters ${k=\left\{ -1,0,+1 \right\}}$ in ${H_0^2 c^{-2}}$ units and the flat $\Lambda$CDM model by fitting them to cosmic chronometers (CC), the Pantheon+ sample of type Ia supernovae (SNe), and standardized quasars (QSOs). We used the \texttt{emcee} code for fitting CC data, a custom Markov Chain Monte Carlo implementation for SNe and QSOs, and Anderson-Darling tests for normality on normalized residuals for model comparison. Best-fit parameters are presented, constrained by data within redshift ranges $z\leq 2$ for CCs, $z\leq 2.3$ for SNe, and $z\leq 7.54$ for QSOs. Coasting models, particularly the flat coasting model, are generally favored over the flat $\Lambda$CDM model. The overfitting of the flat $\Lambda$CDM model to Pantheon+ SNe and the large intrinsic scatter in QSO data suggest a need to refine error estimates in these datasets. We also highlight the seemingly fine-tuned nature of either the CC data or $\Omega_{\mathrm{m},0}$ in the flat $\Lambda$CDM model to an ${H_1=H_0}$ coincidence when fitting ${H(z)=H_1z+H_0}$, a natural feature of coasting models.