LATIS Data Release: $\sim4200$ Spectra of $z \sim 2-3$ Galaxies, Redshifts, and Intergalactic Medium Tomography Maps

TL;DR

The LATIS data release delivers a large, well-documented spectroscopic catalog and the first extensive IGM tomography maps at $z\sim2.5$, enabling high-fidelity studies of galaxies and the intergalactic medium during cosmic noon. It combines $R\sim1000$ IMACS spectra (7408 targets; 5575 high-confidence redshifts, 4176 at $z>1.7$) with Lyα transmission fluctuations in $4.7\times10^5$ pixels to generate Wiener-filtered 3D IGM maps over $1.65\deg^2$ at $z=2.2$–$2.8$ with $4\,h^{-1}$ cMpc voxels, plus mock surveys for calibration. The release includes detailed data reduction upgrades (flux calibration, ADC corrections, sky subtraction), spectral modeling with mean-flux regularization, robust redshift quality flags, and accessible TSR/SSR maps, collectively enabling a broad range of studies on galaxy evolution, environments, and the IGM. Together with public access to the data and ancillary products, LATIS substantially expands the public redshift census at $z\sim2$–$3$ and provides a new, large-scale, high-resolution tracer of large-scale structure in legacy fields.

Abstract

We present the data release of the Ly$α$ Tomography IMACS Survey (LATIS), one of the largest optical spectroscopic surveys of faint high-redshift galaxies. The survey provides 7408 optical spectra of candidate $z \sim 2-3$ galaxies and QSOs in the Canada-France-Hawaii Telescope Legacy Survey D1, D2 (COSMOS), and D4 fields. The $R \sim 1000$ spectra were obtained using the Inamori Magellan Areal Camera and Spectrograph (IMACS) at the Magellan Baade telescope, with typical integrations of 12 hr. From these spectra, we measured 5575 high-confidence spectroscopic redshifts, of which 4176 are at $z > 1.7$, thereby substantially increasing the number of public spectroscopic redshifts at $z \approx 2-3$ in COSMOS and the other survey fields. The data release includes Ly$α$ transmission fluctuations measured in $4.7 \times 10^5$ pixels, which were used to create 3D maps of the intergalactic medium (IGM) transmission spanning 1.65 deg$^2$ and $z = 2.2-2.8$ at a resolution of 4 $h^{-1}$ cMpc. These are the largest such maps to date and provide a novel tracer of large-scale structure in legacy fields. We also provide ancillary data, including mock surveys. The LATIS data will enable a variety of community studies of galaxy evolution, environments, and the IGM around cosmic noon.

Figures (12)

• Figure 1: Maps of the LATIS survey fields. Filled circles show targets that were either confirmed to lie in the targeted redshift range $z=2.2$--2.8, or were used as Ly$\alpha$ forest sight lines to map absorption in that range. Open squares show other galaxies for which high-confidence redshifts were obtained. Red crosses show galaxies for which we were unable to measure a high-confidence redshift. Blue circles show unobserved galaxies in the parent sample to a limit of $r < 24.4$, which were given top priority; many fainter galaxies ($r = 24.4$--24.8) are present in the parent sample and were targeted at lower priority, but these are not shown for clarity. The gray background shows the survey area, comprised of the union of the IMACS footprints with regions around bright stars omitted. (A few galaxies in the southwest of the D1 field lie outside this nominal survey area, because the mask center was shifted after initial observations to capture better guide stars.) The approximate footprints of the COSMOS-Web NIRCam Casey23, 3D-HST Brammer12, CLAMATO DR2 Horowitz22, and VUDS LeFevre15 surveys are shown; the MOSDEF Kriek15 and CANDELS Grogin11Koekemoer11 footprints are similar to 3D-HST. Overlap with other surveys is illustrated in Fig. 1 of Newman20.
• Figure 2: Left panel: The distribution of total integration times for targets with high-confidence redshifts $z > 2$. Right panel: The distribution of the median CNR per spectral pixel for the same targets, evaluated in 40 Å intervals centered on the rest-frame wavelengths indicated in the caption, which were chosen to be free of strong absorption or emission lines.
• Figure 3: Top panel: Average ratio of observed and model spectra for $z=1.8$--3.0 LBGs, in the observed frame. The Ly$\alpha$ forest and strong interstellar absorption lines are excluded from each spectrum. A polynomial fit is shown in blue. Middle panel: Same as the top panel, but with the polynomial fit divided out. A Gaussian model of the systematic residual centered at 4332 Å is overlaid; this was used to correct the flux calibration curve derived from white dwarf standards. Bottom panel: White dwarf standard spectrum demonstrating that the feature arose due to mismatch at H$\gamma$.
• Figure 4: Comparison of LATIS redshifts $z_{\rm LATIS}$ to literature redshifts $z_{\rm lit}$ in the COSMOS compilation of Khostovan25 and VUDS. See the text for discussion of the quality flags used. Circles and stars include LATIS LBGs and QSOs, respectively.
• Figure 5: The distribution of confident ( zqual = 3 or 4) galaxy redshifts measured in each LATIS survey field.