Is Dynamical Dark Energy Necessary? DESI BAO and Modified Recombination

TL;DR

The paper investigates whether the DESI BAO tension with Planck CMB and SN data within ΛCDM can be resolved by altering early-Universe recombination rather than invoking dynamical dark energy. It endorses a four-parameter phenomenological modification to the ionization history, implemented in CAMB and analyzed with Cobaya, showing that altered recombination can achieve a good fit to BAO and CMB while predicting a higher $H_0$ and a larger sound-horizon scale ($r_{ m d}$), thereby partially easing the Hubble tension. When SN data and SH0ES priors are included, the modified recombination scenario remains competitive and is often favored over $w_0w_a$ dynamics, suggesting an early-Universe solution as a viable alternative to late-time dark energy evolution. The results emphasize the potential role of early-Universe physics, including possible primordial magnetic fields or varying constants, in addressing current cosmological tensions, while highlighting the need for future data to robustly differentiate between these possibilities.

Abstract

Recent measurements of baryon acoustic oscillations (BAO) by the Dark Energy Spectroscopic Instrument (DESI) exhibit a mild-to-moderate tension with cosmic microwave background (CMB) and Type Ia supernova (SN) observations when interpreted within the $Λ$CDM framework. This discrepancy has been cited as evidence for dynamical dark energy (DDE). Given the profound implications of DDE for fundamental physics, we explore whether the tension can instead be resolved by modifying the physics of recombination. We find that a phenomenological model of modified recombination can effectively reconcile the BAO and CMB datasets and, unlike DDE, also predicts a higher Hubble constant $H_0$, thereby partially alleviating the Hubble tension. A global fit to BAO, CMB, and calibrated SN data favors modified recombination over DDE.

Figures (2)

• Figure 1: The 68% and 95% confidence level (CL) contours and posterior distributions of $\Omega_{\rm m}$ and $r_{\rm d}h$ from DESI DR2 BAO (DESI2), and combination of DESI2 with the CMB acoustic scale angle $\theta_{\rm CMB}$. Contours for the $\Lambda$CDM model fit to the Planck data and to the combination of ACT DR6 CMB anisotropies and CMB lensing (ACT-L) spectra are also shown for comparison.
• Figure 2: The 68% and 95% CL contours for the $H_0$, $\Omega_{\rm m}$, and $S_8$ along with their posterior distributions for the $\Lambda$CDM, $w_0w_a$, and 4-parameter models fit to combinations of Planck (PL), DESI2, and PP (left), and PP with the SN brightness magnitude prior from SH0ES (PP+$M_b$) (right). For reference, the blue and grey vertical bands show the 68% CL intervals corresponding to $H_0 = 70.4 \pm 1.9$ km/s/Mpc from CCHP, and $H_0 = 73.04 \pm 1.04$ km/s/Mpc from SH0ES, respectively.