Three-form dark energy: constraints and multi-probe comparison with $\Lambda$CDM

Technical Summary: Three-form Dark Energy: Constraints and Multi-probe Comparison with $\Lambda$CDM

Problem Statement
The standard $\Lambda$CDM cosmological model, while successful in describing the cosmic expansion history and large-scale structure formation, faces persistent observational tensions, most notably the discrepancy in the Hubble parameter ($H_0$) between early-universe inferences (CMB) and late-universe distance-ladder measurements. Additionally, the cosmological constant ($\Lambda$) suffers from theoretical challenges, including the fine-tuning and coincidence problems. These issues have motivated the exploration of dynamical dark energy (DE) models and alternative gravity theories. Three-form fields, which arise naturally in higher-dimensional theories such as string theory and supergravity, offer a theoretically well-motivated framework for DE. They can mimic a cosmological constant or exhibit dynamical behaviors, including phantom regimes ($w < -1$), depending on the potential. While previous theoretical studies have explored three-form cosmologies, observational constraints on these models, particularly those driven by a Gaussian potential, remain limited. This work aims to provide a comprehensive observational analysis of a minimally coupled three-form dark energy model with a Gaussian potential, comparing its viability against $\Lambda$CDM using current multi-probe cosmological data.

Methodology
The authors investigate a minimally coupled three-form field $A_{\mu\nu\rho}$ with a self-interaction Gaussian potential $V(\chi) = V_* \exp(-\xi \kappa^2 \chi^2 / 6)$ in a spatially flat FLRW universe. The dynamics are governed by the field equation of motion and the Friedmann equations, where the field exhibits distinct "small-field" and "large-field" branches. The study employs a Bayesian Markov-chain Monte Carlo (MCMC) framework using the Cobaya package to constrain the model parameters against six distinct combinations of cosmological datasets:

1. Early-time data: CMB shift parameters and DESI DR2 Baryon Acoustic Oscillation (BAO) measurements.
2. Late-time data: Pantheon+ Type Ia supernovae (with SH0ES calibration, denoted PPS, and without, denoted Pan), Cosmic Chronometers (CC), and Gamma-Ray Bursts (GRB).
3. Combinations: Six specific dataset groupings were analyzed, including isolated early-time, isolated late-time, and combined sets (e.g., CMB+DESIBAO2+PPS and CMB+DESIBAO2+CC+GRB+Pan).

The analysis involves 7 free parameters for the three-form model (including $H_0$, $\Omega_{m0}$, $\Omega_{b0}$, and four three-form specific parameters defining the potential shape and initial conditions) compared to 3 for $\Lambda$CDM. Model comparison is performed using multiple information criteria: the corrected Akaike Information Criterion (AICC), Bayesian Information Criterion (BIC), Deviance Information Criterion (DIC), Widely Applicable Information Criterion (WAIC), and Bayesian Evidence ($\ln B$). Additionally, tension statistics (Goodness of Fit and Suspiciousness) are calculated to quantify the consistency between early- and late-time datasets.

Key Results

• Parameter Constraints: The three-form model is consistent with current observations. The parameter $\log_{10} \xi$ (controlling the potential width) is generally consistent with the lower boundary of the prior ($\xi \to 0$, the $\Lambda$CDM limit), though some combined datasets show a preference for non-zero values. The initial field amplitude parameter is constrained by early-time data (CMB+BAO), which limits the redshift at which the field can deviate significantly from $\Lambda$CDM behavior.
• Equation of State Dynamics: The reconstructed dark energy equation of state ($w_\chi$) exhibits a characteristic evolution: it remains close to $-1$ in the asymptotic past and future, approaches a Little Sibling of the Big Rip (LSBR) attractor at late times, and displays a distinct "phantom dip" ($w_\chi < -1$) at intermediate redshifts. The timing of this dip is determined by the initial field amplitude.
• Model Comparison:
  • For datasets restricted to either early-time or late-time observations alone, information criteria generally favor $\Lambda$CDM or yield inconclusive results, as the additional degrees of freedom in the three-form model are not strongly constrained and add complexity without sufficient fit improvement.
  • For the combination of early-time and late-time data with high tension (specifically CMB+DESIBAO2+PPS), the three-form model shows a mild preference over $\Lambda$CDM according to AICC, DIC, WAIC, and Bayesian Evidence. This suggests the model's extra degrees of freedom help accommodate the conflicting data origins.
  • The BIC remains positive for all combinations, favoring $\Lambda$CDM due to the penalty term for the number of parameters, particularly in large datasets.
• Hubble Tension: While the three-form model allows for slightly higher values of $H_0$ in combinations including CMB data, it does not significantly alleviate the Hubble tension between CMB+BAO and PPS. The tension remains at the $\sim 5\sigma$ level. The reduction in raw tension statistics (GoF and Suspiciousness) is offset by the effective Bayesian model dimension, which increases when the joint fit must accommodate the tension, thereby steepening the mapping to $\sigma$ values.

Significance and Claims
The paper concludes that the Gaussian three-form dark energy model provides a viable and competitive description of the Universe's expansion history. It offers a well-motivated alternative to $\Lambda$CDM, particularly in scenarios where early- and late-time datasets are combined, as the phantom-like behavior at intermediate redshifts allows for a better accommodation of observational tensions than the standard model. However, the authors emphasize that the statistical preference is modest and not decisive; the model does not resolve the Hubble tension on its own. The distinctive observational signature of this model is the phantom dip in the equation of state, the timing of which is constrained by early-universe data. The study suggests that future cosmological surveys could further test these dynamical features, but currently, the model remains a theoretically rich, observationally consistent, yet not yet decisively preferred extension of the standard cosmological paradigm.