Nonparametric late-time expansion history reconstruction and implications for the Hubble tension in light of recent DESI and type Ia supernovae data
Jun-Qian Jiang, Davide Pedrotti, Simony Santos da Costa, Sunny, Vagnozzi

TL;DR
This paper nonparametrically reconstructs the late-time expansion history of the universe using recent DESI BAO and supernova data, revealing features that challenge standard models and reinforce the sound horizon tension, with implications for dark energy theories.
Contribution
It introduces a nonparametric reconstruction method combining Gaussian Processes and cubic splines to analyze expansion history, identifying features inconsistent with simple dark energy models.
Findings
Deviations from $\\Lambda$CDM are constrained to be less than 10% at $z \lesssim 2$.
Identifies stable features at $z \sim 0.5$ and $z \sim 0.9$ in the expansion rate.
Reinforces the high-significance tension in the sound horizon scale, suggesting new physics.
Abstract
We nonparametrically reconstruct the late-time expansion history in light of the latest Baryon Acoustic Oscillation (BAO) measurements from DESI combined with various Type Ia Supernovae (SNeIa) catalogs, using interpolation through piece-wise natural cubic splines, and a reconstruction procedure based on Gaussian Processes (GPs). Applied to DESI BAO and PantheonPlus SNeIa data, both methods indicate that deviations from a reference CDM model in the unnormalized expansion rate are constrained to be , but also consistently identify two features in : a bump at , and a depression at , which cannot be simultaneously captured by a CDM fit. These features, which are stable against assumptions regarding spatial curvature, interpolation knots, and GP kernel, disappear if one adopts the older SDSS BAO measurements…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
