DESI 2024 IV: Baryon Acoustic Oscillations from the Lyman Alpha Forest

TL;DR

This paper reports the first-year DESI measurement of Baryon Acoustic Oscillations from the Lyman-α forest, using over 420,000 Lyα spectra cross-correlated with 700,000 quasars. It introduces a blinded analysis workflow, a distortion-matrix approach to account for continuum-fitting effects, and a comprehensive treatment of contaminants (metals, HCDs, BALs) with cross-covariance between Lyα auto- and Lyα–quasar cross-correlations. The joint analysis yields precise BAO parameters at z_eff = 2.33, translating to constraints on D_H(z)/r_d and D_M(z)/r_d with ~2% precision, and provides robust cosmological distance measurements that complement lower-redshift DESI BAO results. Extensive validation with mocks and data splits demonstrates the robustness of the BAO inference and clarifies systematic contributions, paving the way for future DESI data to further tighten cosmological constraints.

Abstract

We present the measurement of Baryon Acoustic Oscillations (BAO) from the Lyman-$α$ (Ly$α$) forest of high-redshift quasars with the first-year dataset of the Dark Energy Spectroscopic Instrument (DESI). Our analysis uses over $420\,000$ Ly$α$ forest spectra and their correlation with the spatial distribution of more than $700\,000$ quasars. An essential facet of this work is the development of a new analysis methodology on a blinded dataset. We conducted rigorous tests using synthetic data to ensure the reliability of our methodology and findings before unblinding. Additionally, we conducted multiple data splits to assess the consistency of the results and scrutinized various analysis approaches to confirm their robustness. For a given value of the sound horizon ($r_d$), we measure the expansion at $z_{\rm eff}=2.33$ with 2\% precision, $H(z_{\rm eff}) = (239.2 \pm 4.8) (147.09~{\rm Mpc} /r_d)$ km/s/Mpc. Similarly, we present a 2.4\% measurement of the transverse comoving distance to the same redshift, $D_M(z_{\rm eff}) = (5.84 \pm 0.14) (r_d/147.09~{\rm Mpc})$ Gpc. Together with other DESI BAO measurements at lower redshifts, these results are used in a companion paper to constrain cosmological parameters.

Figures (17)

• Figure 1: Left: Expected final DESI (blue) and SDSS-DR16 footprint (red) together with the spatial distribution of DESI DR1 observed quasars (green). For reference we also show the Galactic plane (solid black) and the Ecliptic plane (dotted black). Right: Number of observations for Ly$\alpha$ quasars in the DESI DR1 sample.
• Figure 2: Left: Redshift distribution of quasars in DESI DR1 (orange), compared to the distribution in SDSS DR16 (green) and to the distribution of Ly$\alpha$ pixels (blue, divided by 200). Right: Contribution of different redshifts to the four measured correlation functions. In particular, we show the sum of weights used in \ref{['eqn:auto_corr', 'eqn:cross_corr']} as a function of redshift, for large transverse separations (to reduce the biasing effect from the clustering of background quasars) and $r_\parallel=0$ (the contribution varies as a function of $r_\parallel$, especially for quasar cross-correlations). As described in \ref{['sec:correlations']}, we measure Ly$\alpha$ correlations in two different rest-frame wavelength regions of the quasar spectra (A and B).
• Figure 3: QSO spectrum from the first year of DESI data at redshift $z=3.14$ (TargetID = 39627581225438176). The region B is highlighted in purple. The region A is highlighted in indigo. The CIV and CIII regions are highlighted in various shades of green. While there is almost no CIII absorption, the CIV absorption spans leftward of the CIV doublet, contaminating the Ly$\alpha$ regions A and B. The Ly$\alpha$ absorption extends into region B. For better visualization, we have chosen a relatively high signal-to-noise spectrum and we have smoothed it by averaging the pixels in groups of 25.
• Figure 4: Measured Ly$\alpha$ auto-correlation when using pixels from region A (top, colored markers) and when correlating pixels from region A with pixels from region B (bottom), along with the best fit model (solid black curves), described in \ref{['sec:model']}. The different colors and markers correspond to different orientations with respect to the line-of-sight, with blue correlations being close to the line-of-sight $0.95<\mu<1$. The dotted curves show the best fit model with additive polynomial corrections (see \ref{['subsec:broadband-validation']}).
• Figure 5: Measured Ly$\alpha\times$QSO cross-correlation functions in region A (top, colored markers) and region B (bottom) along with the best fit model (solid black curves), described in \ref{['sec:model']}. The different colors and markers correspond to different orientations with respect to the line-of-sight, with blue correlations being close to the line-of-sight $0.95<\mu<1$. The dotted curves show the best fit model with additive polynomial corrections (see \ref{['subsec:broadband-validation']}).