Hierarchical Gaussian-process reconstruction of late-time expansion with DESI BAO and compressed Planck priors
Abstract
We reconstruct the late-time expansion history with a hierarchical Gaussian process (GP) that co-samples the kernel hyperparameters (σf,l) together with (Hzero,Om,Ok,ωb h2). The data are 37 cosmic-chronometer H(z) points, DESI DR2 BAO, Planck 2018 compressed distance priors, and Pantheon+ supernovae, coupled by a Monte-Carlo effective likelihood. DESI's headline Chevallier--Polarski--Linder (CPL) preference at $2.8−−4.2σ$ uses the full Planck likelihood; we do not reproduce or refute that result. On this compressed-CMB pipeline the baseline posterior is w(z≃ 0)=−0.80+0.26−0.23 (68\%~C.L.), less than 1σ from w=−1, with l=3.79+0.81−1.12. A CPL fit on the same data improves nested lcdm by only Δχ2∼ 1; Akaike's criterion prefers nested lcdm (ΔAIC≃ 3), while BIC is quoted only as an indicative check for correlated supernovae. The reconstructed fDE=1 lies inside the 68\% band at every grid point with z≤ 1.6. Ablations that fix (σf,l), drop SN or LRG1/2 BAO, replace the radial-basis kernel by a Matérn family, swap the SN catalogue (Union3, DES-Y5, DES-Dovekie), or nest a residual GP around lcdm leave the median w(0) within ≲ 1σ of −1.
Related
- Adaptive proposals for nonparametric equation-of-state inference with RIFT: iterative Gaussian-process refinement in a compressed sequence representation
- Enhancing dark siren cosmology via Gaussian process reconstruction of incomplete galaxy catalogs
- A gaussian-process reconstruction of gravitational-wave friction from dark standard sirens
- Gaussian process inference of stochastic magneto-active dynamics and viscosity in Swift J1727.8-1613
- A Spatio-Temporal Self-Propagating Log-Gaussian Cox-Hawkes Process for Star Formation Modelling
BibTeX
@misc{li2026hierarchical,
title = {{Hierarchical Gaussian-process reconstruction of late-time expansion with DESI BAO and compressed Planck priors}},
author = {Enkun Li and Yu-Hao Mu and Zhihuan Zhou},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2607.23593},
eprint = {2607.23593},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2607.23593},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Hierarchical Gaussian-process reconstruction of late-time expansion with DESI BAO and compressed Planck priors
AU - Enkun Li
AU - Yu-Hao Mu
AU - Zhihuan Zhou
PY - 2026
DA - 2026/09/19
PB - arXiv
DO - 10.48550/arxiv.2607.23593
UR - https://arxiv.org/abs/2607.23593
AB - We reconstruct the late-time expansion history with a hierarchical Gaussian process (GP) that co-samples the kernel hyperparameters (σ_f,l) together with (Hzero,Om,Ok,ω_b h^2). The data are 37 cosmic-chronometer H(z) points, DESI DR2 BAO, Planck 2018 compressed distance priors, and Pantheon+ supernovae, coupled by a Monte-Carlo effective likelihood. DESI's headline Chevallier--Polarski--Linder (CPL) preference at $2.8−−4.2σ$ uses the full Planck likelihood; we do not reproduce or refute that result. On this compressed-CMB pipeline the baseline posterior is w(z≃ 0)=−0.80^+0.26_−0.23 (68\%~C.L.), less than 1σ from w=−1, with l=3.79^+0.81_−1.12. A CPL fit on the same data improves nested lcdm by only Δχ^2∼ 1; Akaike's criterion prefers nested lcdm (ΔAIC≃ 3), while BIC is quoted only as an indicative check for correlated supernovae. The reconstructed f_DE=1 lies inside the 68\% band at every grid point with z≤ 1.6. Ablations that fix (σ_f,l), drop SN or LRG1/2 BAO, replace the radial-basis kernel by a Matérn family, swap the SN catalogue (Union3, DES-Y5, DES-Dovekie), or nest a residual GP around lcdm leave the median w(0) within ≲ 1σ of −1.
ER -