Testing Quintessence Axion Dark Energy with Recent Cosmological Results

TL;DR

This paper tests a quintessence axion model for dynamical dark energy against DESI DR2 BAO, Planck compressed CMB, and SNe Ia data. The axion potential is a cosine form $V(A)=\frac{V_0}{2}\left[1-\cos\left(\frac{A}{F_A}\right)\right]$, with mass $m_A^2=\frac{V_0}{2F_A^2}$, and the authors perform MCMC analyses over the parameters via an $\epsilon$-based sampling to avoid biases from tiny initial displacements. They find a mild preference for sub-Planckian $F_A$ (and hence larger $m_A/H_0$), while a trans-Planckian regime remains allowed; current data constrain the distance–redshift relation rather than $F_A$ itself, and the best-fit often correlates $F_A/M_P$ with $m_A/H_0$ along the approximate relation $F_A/M_P \sim \frac{3}{2}(1-\Omega_m)\frac{H_0}{m_A}$. The analysis with cosmic birefringence data favors an EM anomaly coefficient $c_\gamma=\mathcal{O}(1)$ for certain ranges, linking dark energy dynamics to parity-violating photon couplings and offering a testable signature for future polarization measurements.

Abstract

We investigate a quintessence axion model for dynamical dark energy, motivated in part by recent results from the Baryon Acoustic Oscillation (BAO) measurements of the Dark Energy Spectroscopic Instrument (DESI) combined with the cosmic microwave background anisotropies and the latest Type Ia supernovae (SNe Ia) data. By carefully treating the initial conditions and parameter sampling, we identify a preferred parameter space featuring a sub-Planckian axion decay constant and a relatively large axion mass, which naturally avoids the quality problem and remains consistent with the perturbative string conjecture. Our parameter scan also uncovers a trans-Planckian regime of theoretical interest, which is only mildly disfavored even by the strongest constraint. Finally, we discuss the possible connection between this model and the recently reported non-zero rotation of the CMB linear polarization angle, emphasizing the broader cosmological implications and the promising prospects for testing this scenario. We show that an $\mathcal{O}(1)$ electromagnetic anomaly coefficient is preferred by the strongest constraint, which is in full agreement with the minimal quintessence axion model.

Figures (4)

• Figure 1: Constraints in the $m_A / H_0$ versus $\epsilon$ plane from our fiducial analysis using the $\epsilon$-sampling method. This sampling strategy ensures adequate exploration of the small-$m_A$ region, as illustrated in the plot.
• Figure 2: Constraints on the $m_A/H_0$ versus $F_A/M_P$ plane (upper panel) and the $\delta_i$ versus $F_A/M_P$ plane (lower panel) from our fiducial analysis. Upper: The dotted curve shows the relation in Eq. \ref{['eq:fit']}. Lower: The dotted curve shows the relation in Eq. \ref{['eq:deltai-kf-estimate']}, when the equal sign is restored. The dotted-dashed line shows $\delta_i=\pi$, which corresponds to the limit where the quintessence axion potential reduces to a quadratic potential.
• Figure 3: Results using a uniform $\delta_i$ sampling with $0.01<\delta_i<\pi$. The lower limit of $m_A$ shown here is only set by the parameter sampling method.
• Figure 4: Constraints on the $c_\gamma$ versus $F_A/M_P$ plane when including CB in the analysis. Notably, for DESI + CMB + DESY5 + CB, an electromagnetic anomaly coefficient $c_\gamma$ of $\mathcal{O}(1)$ is preferred, which is consistent with the prediction of the minimal model.