arXiv · 2611.18558
Reconstruction of Dynamical Dark Energy and Gravity from latest cosmological data
Abstract
Recent cosmological observations have renewed interest in non-standard late-time evolution. We investigate whether departures from the standard expansion history are accompanied by modifications of gravitational dynamics through a joint, non-parametric reconstruction of the effective dark energy density, $Ω_{\rm X}(a)$, and the gravitational response functions $μ(a)$ and $Σ(a)$. Combining CMB, DESI DR2 BAO, Pantheon+, and large scale structure observations, including DES Year 3 $6\times2$pt data, we perform reconstructions with and without a theory motivated correlation prior derived from stable Horndeski models. We find a persistent maximum in $Ω_{\rm X}$ at $z\simeq0.5$--$0.7$, corresponding to a crossing of the effective phantom divide. This feature is already present in the CMB+BAO combination and remains robust when modifications to clustering and lensing are allowed. The reconstructed gravitational response exhibits suggestive enhancements in $μ$ and $Σ$ over approximately the same redshift range, although these departures are less significant. Including LSS information instead favours a reduced lensing response, $Σ<1$, while leaving the background feature largely unchanged. An interpretation within Generalized Brans-Dicke theories illustrates how modifications to the effective gravitational coupling and a scalar-mediated fifth force may contribute to these trends. Our results indicate that current departures from $Λ$CDM are most significant in the expansion history, while leaving open the possibility of correlated modifications to gravitational dynamics that future Stage IV surveys can test.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Maria Luisa Gallavotti, Alessandra Silvestri. 2026-11-13. Reconstruction of Dynamical Dark Energy and Gravity from latest cosmological data. https://arxiv.science/abs/2611.18558
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite arXiv Science by Indiciae: BibTeX · RIS.