Direct cosmological inference from three-dimensional correlations of the Lyman-$α$ forest

Abstract

When performing cosmological inference, standard analyses of the three-dimensional Lyman-$\alpha$ (Ly$\alpha$) correlation functions typically retain only the information carried by the distinct peak produced by baryon acoustic oscillations (BAO). In this work, we investigate whether this compression is sufficient to capture all the relevant cosmological information contained in these correlations. We address this question by performing a direct fit to the full shape of synthetic Ly$\alpha$ auto-correlation functions and their cross-correlations with quasars, including all physical scales without compression, at an effective redshift $z_{\mathrm{eff}} = 2.3$, assuming a DESI-like survey, and by comparing the results to those obtained with the standard BAO-based approach applied to the same data set. Our method yields a $3.5%$ constraint on the matter density, $\Omega_m$, which is approximately three to four times tighter than what can be achieved using BAO information alone. The growth parameter $f\sigma_8(z_{\mathrm{eff}})$ is constrained at the $10%$ level, and the scalar spectral index $n_s$ at the $\sim 3$–$4%$ level. We show that the additional information gained from the direct-fit approach, with the exception of the constraint on $n_s$, can be primarily attributed to sensitivity to the Alcock–Paczyński effect and redshift-space distortion information.

Publication
arXiv