Bimetric gravity improves the fit to DESI BAO and eases the Hubble tension

Marcus Högås, Edvard Mörtsell · 2025

Read Bimetric gravity improves the fit to DESI BAO and eases the Hubble tension by Marcus Högås, Edvard Mörtsell (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-03743.txt.

Introduction

Marcus Högås and Edvard Mörtsell test a modified theory of gravity against new cosmic data in a 2025 paper. The data combine galaxy distance measures, the faint glow left from the early universe, and exploding stars used as distance markers. A live question is whether dark energy changes over time.

One common model lets dark energy vary, and it fits the data somewhat better than a fixed value. But that model worsens the clash between two measures of how fast the universe expands. The authors' bimetric gravity, which uses two linked measures of space, fits better and eases that clash at the same time. The paper makes a case for taking the modified theory seriously as data sharpen.

Catalog record

CREATORMarcus Högås, Edvard Mörtsell
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-03743

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Bimetric gravity improves the fit to DESI BAO and eases the Hubble tension Marcus Högås, Edvard Mörtsell Abstract We investigate whether the latest combination of DESI DR2 baryon acoustic oscillation (BAO) measurements, cosmic microwave background (CMB) data ( Planck 2018 + ACT), and Type Ia supernovae (SNe Ia) compilations (Pantheon+, Union3, and DES Y5) favor a dynamical dark energy component, and explore if such a scenario can simultaneously help resolve the Hubble tension. We contrast two frameworks: the widely used phenomenological w 0 ​ w a w_{0}w_{a} CDM model, and bimetric gravity—a fundamental modification of general relativity that naturally gives rise to phantom dark energy. The w 0 ​ w a w_{0}w_{a} CDM model is moderately preferred over Λ \Lambda CDM, at the 2 2 – 4 ​ σ 4\,\sigma level, when fitting DESI DR2 + CMB + SNe Ia, but it exacerbates the Hubble tension. By comparison, bimetric gravity provides a modest improvement in fit quality, at the 1 ​ σ 1\,\sigma level, but, by inferring H 0 = 69.0 ± 0.4 ​ km / s / Mpc H_{0}=69.0\pm 0.4\,\mathrm{km/s/Mpc} , it partially eases the Hubble tension, from a 5 ​ σ 5\,\sigma discrepancy to a 3.7 ​ σ 3.7\,\sigma tension. Including locally calibrated SNe Ia brings the overall preference for the bimetric model over Λ \Lambda CDM to the 2 ​ σ 2\,\sigma level, comparable to that of the w 0 ​ w a w_{0}w_{a} CDM model when including the local SN Ia calibration. I Introduction The standard cosmological model, built on general relativity and quantum physics, has a remarkably successful history in predicting and explaining cosmological observations. The discovery of cosmic acceleration in the late 1990s led to the reintroduction of the cosmological constant ( \Lambda ) as a key component of this framework Einstein ; Perlmutter et al. 1999 ; Riess et al. 1998 . However, new data hints at an evolving dark energy component. In particular, the Dark Energy Spectroscopic Instrument (DESI) Year 3 data release (DR2)—which provides an unprecedented 3D map of more than 14 million galaxies and quasars—has enabled the most precise BAO measurements to date. These measurements, when combined with complementary data from the CMB and SNe Ia, suggest that the dark energy density may evolve with time Lodha et al. 2025 ; Abdul Karim et al. 2025 . However, the statistical significance of evolving dark energy is still modest—typically \lesssim 4\,\sigma , depending on the dataset combination and parameterization—and should be interpreted with appropriate caution. For instance, systematic uncertainties in the SNe Ia can significantly affect the inferred preference for dynamical dark energy Dhawan et al. 2025 ; Gialamas et al. 2025 ; Huang et al. 2025 ; Afroz and Mukherjee 2025 ; Dhawan and Mörtsell 2025 , possibly due to subtle effects in their cross-calibration Popovic et al. 2025 . On the other hand, a majority of dataset combinations—spanning BAO measurements from DESI and the Sloan Digital Sky Survey (SDSS), CMB data from Planck , the Atacama Cosmology Telescope (ACT), and the South Pole Telescope (SPT), and SNe Ia from Pantheon+, Union3, and the Dark Energy Survey Year 5 (DES Y5) data compilations—do favor a time-evolving dark energy. The strength of this preference varies, and a few specific combinations—such as SDSS BAO with Pantheon+ SNe Ia, or CMB data from SPT—show no significant evidence for dynamical evolution Giarè et al. 2025 ; Giarè 2025 . Much of contemporary literature study phenomenological models of dark energy, with the Chevallier–Polarski–Linder (CPL) parameterization being the most widely used, with the equation of state w_{\mathrm{DE}}(a)=w_{0}+w_{a}(1-a) as a function of the scale factor, a . We will refer to this as the w_{0}w_{a} CDM model. Although computationally efficient, w_{0}w_{a} CDM is not derived from fundamental physics and thus lacks theoretically justified parameter priors. Still, such models can offer guidance on which features valid fundamental theories may need to accommodate. One feature allowed for in w_{0}w_{a} CDM is phantom crossing, where the dark energy equation of state w_{\mathrm{DE}}(a) evolves across the phantom divide w=-1 Özülker et al. 2025 . Realizing such an evolution within a fundamental physics framework such as a scalar field model of dark energy faces theoretical challenges, including instabilities and fine-tuning Carroll et al. 2003 ; Vikman 2005 , although there are exceptions Clifton et al. 2012 ; Hu 2005 ; Ishak 2019 ; Koussour et al. 2023 ; Wolf et al. 2025 . Another possible hint for dynamical dark energy is the Hubble tension—the persistent discrepancy between different estimates of the Hubble constant ( H_{0} ) which sets the present-day expansion rate of the Universe. CMB observations yield a low value, H_{0}=67.4\pm 0.5\,\mathrm{km/s/Mpc} Aghanim et al. 2020a , when interpreted within the \Lambda CDM framework. In contrast, late-Universe measurements, most notably the SH0ES team’s distance ladder estimate, give a significantly higher value, H_{0}=73.0\pm 1.0\,\mathrm{km/s/Mpc} Riess et al. 2022 . This 8\,\% difference exceeds the 5\,\sigma level and has sparked extensive debate in the literature. A possible resolution may be unaccounted-for systematics in the distance ladder, see for example Mortsell et al. 2022a ; Mortsell et al. 2022b ; Högås and Mörtsell 2025 . Another possible solution involves modifying the cosmological expansion history, for example through a dynamical dark energy component that increases the CMB-inferred value of H_{0} . See Refs. Di Valentino et al. 2021 ; Schöneberg et al. 2022 ; Abdalla et al. 2022 for reviews. Taken together, the DESI DR2 results and the Hubble tension present two potential hints of dynamical dark energy. A natural question is whether they point toward a common form of dark energy. For the CPL parameterization, the answer appears to be no: when fitting DESI DR2 jointly with CMB and SNe Ia data, the inferred Hubble constant falls even below the \Lambda CDM value of 67.4\,\mathrm{km/s/Mpc} , thereby worsening the tension. Executive summary. In this work, we show that bimetric gravity—a natural extension of general relativity—is preferred over \Lambda CDM at the 1\,\sigma level by DESI DR2 BAO + CMB + SNe Ia data, while also predicting a higher Hubble constant, H_{0}=69.0\pm 0.4\,\mathrm{km/s/Mpc} , thereby easing the Hubble tension to 3.7\,\sigma from an initial >5\,\sigma discrepancy. Although current data offer intriguing but not yet decisive hints of dynamical dark energy, bimetric gravity stands out as a theoretically motivated scenario. Future observational efforts, including model-independent reductions of the DESI BAO data and targeted searches for bimetric-specific gravitational-wave signatures, will be essential to determine whether these hints reflect genuine new physics beyond \Lambda CDM. II Data description Our analysis incorporates BAO measurements from DESI DR2, derived from observations of galaxies, quasars, and Lyman- \alpha tracers Abdul Karim et al. 2025 . For the CMB, we use a compressed likelihood combining data from the Planck 2018 data release Aghanim et al. 2020b and ACT Data Release 6 Madhavacheril et al. 2024 , as provided in Ref. Bansal and Huterer 2025 . SNe Ia distances and magnitudes are taken from three sources: Pantheon+ Scolnic et al. 2022 ; Brout et al. 2022 , the binned Union3 sample Rubin et al. 2023 , and DES Y5 Abbott et al. 2024 . To minimize systematic effects due to peculiar velocities, we include only supernovae with z>0.023 . 1 1 1 This

Read the complete text