Tantalizing Evidence of Reionization Relics in the eBOSS DR16 Ly$\boldsymbolα$ Forest Correlations: a Preference for Early Reionization

TL;DR

This work tests whether memory of HI reionization leaves detectable broadband imprints in the Ly$\alpha$ forest correlations measured with eBOSS DR16. By incorporating two reionization templates (Yukawa-like and PySR) into the Ly$\alpha$ auto- and cross-correlation models, the authors show that BAO parameters remain nearly unchanged, while full-shape parameters, especially the Ly$\alpha$ bias and RSD, exhibit significant shifts under late reionization scenarios. Across four correlation functions, a consistent, though modest, preference emerges for reionization relics, with early reionization ($z_{\rm re} \sim 8$) providing the best overall fit and an inferred relic amplitude $A_{\rm re} \approx 0.19$. These results imply that eBOSS Ly$\alpha$ data contains a memory of reionization and highlight the importance of modeling broadband reionization effects in future Ly$\alpha$-based cosmological analyses, while noting caveats from covariance and HeII/X-ray preheating uncertainties.

Abstract

Cosmic reionization of HI leaves enduring relics in the post-reionization intergalactic medium, potentially influencing the Lyman-$α$ (Ly$α$) forest down to redshifts as low as $z \approx 2$, which is the so-called ''memory of reionization'' effect. Here, we re-analyze the baryonic acoustic oscillation (BAO) measurements from Ly$α$ absorption and quasar correlations using data from the extended Baryonic Oscillation Spectroscopic Survey (eBOSS) Data Release 16 (DR16), incorporating for the first time the memory of reionization in the Ly$α$ forest. Three distinct scenarios of reionization timeline are considered in our analyses. We find that the recovered BAO parameters ($α_\parallel$, $α_\perp$) remain consistent with the original eBOSS DR16 analysis. However, models incorporating reionization relics provide a better fit to the data, with a tantalizing preference for early reionization, consistent with recent findings from the James Webb Space Telescope. Furthermore, the inclusion of reionization relics significantly impacts the non-BAO parameters. For instance, we report deviations of up to $3σ$ in the Ly$α$ redshift-space distortion parameter and $\sim7σ$ in the linear Ly$α$ bias for the late reionization scenario. Our findings suggest that the eBOSS Ly$α$ data is more accurately described by models that incorporate a broadband enhancement to the Ly$α$ forest power spectrum, highlighting the importance of accounting for reionization relics in cosmological analyses.

Figures (10)

• Figure 1: Wedge plot of the Ly$\alpha$(Ly$\alpha$) auto correlation function for $0.00<\mu<0.50$. The green points represent the eBOSS DR16 data. The black solid curve illustrates the auto-correlation function without any other broadband effects, highlighting the contribution of the BAO peak. Each colored dashed curve represents the result of adding one of the several broadband components of the model, including the Arinyo non-linear correction (dashed blue, §\ref{['ssec:nl']}), high column density systems (dashed purple, §\ref{['ssec:hcd']}), HI reionization relics (dashed yellow, §\ref{['ssec:reio']}), and metal contamination (dashed red, §\ref{['ssec:metal']}). The gray solid curve shows the result of including them all together. The curves in this figure do not correspond to any best-fit because all the free parameters are fixed to make the comparison between different broadband components fair and clear. The reionization relics effect is represented by the PySR late reionization scenario, which gives the strongest deviations relative to the fiducial no reionization model (see §\ref{['ssec:reio']} for a description of the reionization scenarios considered throughout this work).
• Figure 2: Evolution of the HI fraction for the early (green solid), mid (blue dash-dotted), and late (purple dashed) reionization scenarios. For context, we include several astrophysical constraints such as: dark pixel 2023ApJ...942...59J, Ly$\alpha$ emission fraction 2015MNRAS.446..566M, clustering of Ly$\alpha$ emitters 2015MNRAS.453.1843S, Ly$\alpha$ equivalent width of Ly$\alpha$ emitters 2018ApJ...856....2M2019MNRAS.485.3947M2019ApJ...878...12H, the evolution of the galaxy Ly$\alpha$ luminosity function 2021ApJ...919..120M, galaxy damping wings 2025arXiv250111702M, and quasar damping wings 2022MNRAS.512.5390G2024MNRAS.530.3208G2024AA...688L..26S2024ApJ...969..162D. The gray dashed line corresponds to $x_{\rm HI} = 0.5$ while the olive and cyan colored regions indicate the inferred values for the midpoint of reionization obtained from CMB data by 2020AA...635A..99P and 2021MNRAS.507.1072D, respectively.
• Figure 3: Wedges of the Ly$\alpha$(Ly$\alpha$) auto-correlation function of the Ly$\alpha$ forest. The black solid curve corresponds to the best-fit model for the fiducial scenario, which assumes no reionization relics. The red dashed curve represents the early reionization scenario described by the PySR early template, our best performing model. Green points correspond to the measurements from eBOSS DR16. All models include metal contamination, which explains the bumps in the top-left panel at separations of 20, 60, and 111 $h^{-1}$Mpc, arising from correlations with SiII and SiIII 2023JCAP...11..045G.
• Figure 4: Residuals of the 3D Ly$\alpha$(Ly$\alpha$) auto-correlation function. The residuals are defined as the difference between the model and the observed data, normalized by the observational uncertainties, i.e. $(\xi^{\rm model} - \xi^{\rm data}) / \sigma^{\rm data}$. In Figure \ref{['fig:dif_xi_withreio']}, the model incorporates reionization relics, as described by Eq. (\ref{['eq:auto-mem']}), while Figure \ref{['fig:dif_xi_noreio']} represents the fiducial scenario, which excludes reionization effects. While differences between these models appear subtle, they are sufficiently significant to bias the non-BAO parameters in the fit. See Figure \ref{['fig:xi_comp']} for highlight of the difference between the two models.
• Figure 5: Comparison of Figure \ref{['fig:dif_xi_withreio']} and Figure \ref{['fig:dif_xi_noreio']}. For each pixel in Figure \ref{['fig:dif_xi_withreio']} (with reionization, labeled $A$) and Figure \ref{['fig:dif_xi_noreio']} (fiducial, labeled $B$), this plot displays the ratio $A/B$. To improve readability, we subtract 1 from the ratios, centering the values around zero. Around the typical BAO scale of 100 $h^{-1}$Mpc, an arc of lighter pixels is evident, indicating that the impact of reionization at the BAO peak is negligible.