Dynamical dark energy from DESI DR2 and DES Year 6 BAO, supernovae, and cosmic chronometers: a 3σ preference without decisive Bayesian evidence

Abstract

Baryon acoustic oscillation (BAO) measurements from the Dark Energy Spectroscopic Instrument (DESI) and the Dark Energy Survey (DES) have suggested a preference for dynamical dark energy with $w_0\gt -1$ and $w_a\lt 0$. I derive cosmological constraints in the $\Lambda$ cold dark matter ($\Lambda$CDM), wCDM, and $w_0w_a$CDM models from the DESI Data Release 2 (DR2) and DES Year 6 BAO measurements, combined with the full Planck cosmic microwave background (CMB) likelihood, two recent Type Ia supernova compilations (Union3.1 and DES-Dovekie), and 38 cosmic chronometer $H(z)$ measurements that provide a largely model-independent probe of the late-time expansion. For $\Lambda$CDM, the full combination yields $H_{0}=68.41 \pm 0.30 \, \mathrm{km\, s^{-1}\, Mpc^{-1}}$, $\Omega _{\rm m}=0.301 \pm 0.004$, and $r_{\mathrm{drag}}=147.55 \pm 0.20 \, \mathrm{Mpc}$. In the wCDM model, the equation of state is consistent with a cosmological constant once the CMB is included ($w=-0.992 \pm 0.028$). In the dynamical $w_0w_a$CDM model, the data prefer an evolving equation of state, $w_0=-0.740 \pm 0.082$ and $w_a=-0.893 \pm 0.285$, deviating from $(w_0,w_a)=(-1,0)$ at the $3\sigma$ level for both supernova samples, in agreement with recent DESI DR2 analyses. A Bayesian comparison qualifies this: the fit improves over $\Lambda$CDM ($\Delta \chi ^2\simeq -12$), yet the Bayes factor between $w_0w_a$CDM and $\Lambda$CDM is inconclusive ($\mathrm{ln}\, B\simeq -0.5$), and wCDM is disfavoured ($\mathrm{ln}\, B\simeq -4$), leaving a clear preference only for the evolving over the constant equation of state ($\mathrm{ln}\, B\simeq +3.6$). The data thus support a mild but persistent fit-level preference for an evolving equation of state, robust to the supernova compilation, that is not yet matched by decisive Bayesian evidence.