Probing the Evolution of Dark Energy: A Joint Analysis of DESI DR2, Pantheon+, and Cosmic Chronometers

Abstract: We investigate several phenomenological dark energy parameterizations using a joint analysis of late-time cosmological observations, including cosmic-chromatometer measurements of the Hubble parameter, DESI DR2 baryon acoustic oscillation data, and the Pantheon+ Type Ia supernova sample. Our results show that allowing for a time-varying dark energy equation of state significantly improves the overall fit compared to $Λ$CDM. The present-day equation-of-state parameter departs from the standard cosmological constant value. In contrast, the evolution parameter in two-parameter models tends to be negative, indicating a possible time dependence of dark energy. However, the constraints on the evolution remain moderate, and current data cannot clearly distinguish the specific functional form of dark energy. Model comparison using information criteria suggests that dynamical dark energy models are favored over $Λ$CDM, with the most straightforward one-parameter extension emerging as the most parsimonious scenario. These findings indicate a mild preference for dark energy evolution, though future high-precision observations will be required for definitive conclusions.

• The paper finds that dynamical dark energy models, especially w0CDM, significantly improve data fit with a Δχ² > 15 over ΛCDM.
• The analysis employs multiple parameterizations using MCMC to jointly analyze DESI DR2, Pantheon+, and cosmic chronometer measurements.
• Implications include a 4σ deviation from w = -1 and modest Hubble tension mitigation, underscoring the importance of evolving dark energy.

Probing Dynamical Dark Energy with DESI DR2, Pantheon+, and Cosmic Chronometers

Motivation and Context

This study delivers a joint, likelihood-based analysis of current late-time cosmological datasets—specifically, DESI DR2 BAO, Pantheon+ Type Ia supernovae, and cosmic chronometer Hubble parameter measurements—to probe the evolution of the dark energy equation-of-state (EoS). The analysis is motivated by persistent theoretical and empirical shortcomings of the $\Lambda$CDM model, including the cosmological constant problem and the statistically significant Hubble tension. Precise BAO and SNe Ia datasets, particularly from DESI DR2 and Pantheon+, now provide an opportunity to robustly evaluate phenomenological extensions to $\Lambda$CDM.

Methodological Framework

The work systematically implements parameterizations of $w_{DE}(z)$, including $w_0$CDM (constant, free $w_0$), Chevallier-Polarski-Linder (CPL), Barboza-Alcaniz (BA), Jassal-Bagla-Padmanabhan (JBP), logarithmic, and exponential forms. All scenarios assume a spatially flat FLRW metric, neglect radiation energy density at late times, and numerically solve for $H(z)$ using MCMC (emcee implementation), leveraging the combined likelihood across all datasets.

The model selection leverages both Akaike and Bayesian Information Criteria (AIC/BIC), providing penalization for parameter number and establishing statistical parsimony. The robustness of parameter estimation is ensured by combining SNe Ia, BAO, and cosmic chronometers, which probe complementary redshift regimes.

Figure 1: Marginalized posterior constraints for the $w_0$CDM parameterization, displaying one- and two-dimensional confidence contours for $H_0$, $\Omega_m$, and $w_0$.

Principal Results

The standard $\Lambda$0CDM ($\Lambda$1) fit yields a best-fit $\Lambda$2 km s$\Lambda$3 Mpc$\Lambda$4 and $\Lambda$5. Introduction of dark energy dynamics leads to a substantial reduction in $\Lambda$6 by $\Lambda$7 across all tested extensions, with no significant difference in $\Lambda$8 among the various dynamical EoS parameterizations.

The $\Lambda$9CDM model yields $w_{DE}(z)$0, a statistically significant $w_{DE}(z)$1 deviation from the cosmological constant value. Corresponding $w_{DE}(z)$2 is moderately reduced to $w_{DE}(z)$3 km s$w_{DE}(z)$4 Mpc$w_{DE}(z)$5, addressing a fraction of the Hubble tension, though not eliminating it. Two-parameter models (CPL, BA, JBP, logarithmic, exponential) constrain $w_{DE}(z)$6, with associated evolution parameters $w_{DE}(z)$7 to $w_{DE}(z)$8. While $w_{DE}(z)$9 is negative in all explored cases, uncertainties remain such that these models provide only a mild indication of time-evolving dark energy.

AIC differences ($w_0$0) and favorable BIC for dynamical models indicate strong evidence against $w_0$1CDM in favor of a dynamical $w_0$2. The simplest one-parameter $w_0$3CDM is statically preferred due to its economy, while two-parameter forms are competitive but not distinguishable given current data precision.

Theoretical and Practical Implications

The analysis robustly disfavors $w_0$4 at high significance in combined late-Universe probes, implying either a breakdown of the cosmological constant hypothesis or the presence of systematic effects yet unidentified in these datasets. All dark energy parameterizations with $w_0$5 at $w_0$6 and $w_0$7 suggest a scenario where dark energy density becomes less negative or evolves away from vacuum energy dominance at higher $w_0$8.

From a model-selection perspective, these results lend substantial weight to the investigation of dynamical dark energy phenomenology in late-time cosmology, but current observational precision and degeneracy among parameterizations prohibit distinguishing between different functional forms. The findings underscore the necessity for next-generation datasets with tighter systematics, increased redshift reach, and higher precision on Hubble measurements.

Prospects for Future Exploration

While current CMB-independent late-time constraints signal a mild-to-moderate preference for dark energy evolution, future wide-field BAO, cosmic chronometer, and high-$w_0$9 SNe surveys will be required to distinguish between classes of $w_0$0 models and to statistically confirm or refute the deviation from $w_0$1CDM. In particular, further cross-correlation with early Universe data and careful consideration of cross-survey systematics remain vital for future cosmological inference.

Conclusion

The joint analysis of DESI DR2, Pantheon+, and cosmic chronometer datasets reveals that models of time-varying dark energy are favored over $w_0$2CDM by current late-time cosmological data. The $w_0$3CDM extension, with $w_0$4, provides the best-fit, most parsimonious scenario, though all dynamical models are statistically competitive. Despite this, dataset precision and degeneracy prevent discriminatory power regarding the functional form of $w_0$5. Forthcoming advancements in late-Universe probes are required for decisive statements on the underlying physics of dark energy.

---

Reference: "Probing the Evolution of Dark Energy: A Joint Analysis of DESI DR2, Pantheon+, and Cosmic Chronometers" (2604.05849)