Bayesian Component Separation for DESI LAE Automated Spectroscopic Redshifts and Photometric Targeting Facilities: Mayall (DESI), Blanco (DECam)

Ana Sofía M. Uzsoy, Andrew K. Saydjari, Arjun Dey, Anand Raichoor, Douglas P. Finkbeiner, Eric Gawiser, Kyoung-Soo Lee, Steven Ahlen, Davide Bianchi, David Brooks, Todd Claybaugh, Andrei Cuceu, Axel de la Macorra, Peter Doel, Andreu Font-Ribera, Jaime E. Forero-Romero, Enrique Gaztañaga, Satya Gontc · 2025

Read Bayesian Component Separation for DESI LAE Automated Spectroscopic Redshifts and Photometric Targeting Facilities: Mayall (DESI), Blanco (DECam) by Ana Sofía M. Uzsoy, Andrew K. Saydjari, Arjun Dey, Anand Raichoor, Douglas P. Finkbeiner, Eric Gawiser, Kyoung-Soo Lee, Steven Ahlen, Davide Bianchi, David Brooks, Todd Claybaugh, Andrei Cuceu, Axel de la Macorra, Peter Doel, Andreu Font-Ribera, Jaime E. Forero-Romero, Enrique Gaztañaga, Satya Gontc (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-06870.txt.

Introduction

Ana Sofía M. Uzsoy and a large team of co-authors automate redshift finding for faint galaxies. Their 2025 paper targets Lyman-alpha emitters in DESI spectra. These distant galaxies mark the early universe for cosmology. Sky glow lines, though, can fool automatic redshift tools into errors.

The team builds a Bayesian method that splits each spectrum into parts. Sky leftovers, the galaxy signal, and other residue are inferred jointly. A data-driven prior comes from spectra that humans already inspected. Redshift then follows while sky effects are averaged out. The method steadies automatic classification for survey-scale data. It supports both redshift catalogs and planning of future targets.

Catalog record

CREATORAna Sofía M. Uzsoy, Andrew K. Saydjari, Arjun Dey, Anand Raichoor, Douglas P. Finkbeiner, Eric Gawiser, Kyoung-Soo Lee, Steven Ahlen, Davide Bianchi, David Brooks, Todd Claybaugh, Andrei Cuceu, Axel de la Macorra, Peter Doel, Andreu Font-Ribera, Jaime E. Forero-Romero, Enrique Gaztañaga, Satya Gontc
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE61 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-06870

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Bayesian Component Separation for DESI LAE Automated Spectroscopic Redshifts and Photometric Targeting Facilities: Mayall (DESI), Blanco (DECam) Ana Sofía M. Uzsoy, Andrew K. Saydjari, Arjun Dey, Anand Raichoor, Douglas P. Finkbeiner, Eric Gawiser, Kyoung-Soo Lee, Steven Ahlen, Davide Bianchi, David Brooks, Todd Claybaugh, Andrei Cuceu, Axel de la Macorra, Peter Doel, Andreu Font-Ribera, Jaime E. Forero-Romero, Enrique Gaztañaga, Satya Gontcho A Gontcho, Gaston Gutierrez, Mustapha Ishak, Robert Kehoe, David Kirkby, Anthony Kremin, Martin Landriau, Laurent Le Guillou, Aaron Meisner, Ramon Miquel, John Moustakas, Nathalie Palanque-Delabrouille, Francisco Prada, Ignasi Pérez-Ràfols, Graziano Rossi, Eusebio Sanchez, David Schlegel, Michael Schubnell, Hee-Jong Seo, David Sprayberry, Gregory Tarlé, Benjamin Alan Weaver, Hu Zou Abstract Lyman Alpha Emitters (LAEs) are valuable high-redshift cosmological probes traditionally identified using specialized narrow-band photometric surveys. In ground-based spectroscopy, it can be difficult to distinguish the sharp LAE peak from residual sky emission lines using automated methods, leading to misclassified redshifts. We present a Bayesian spectral component separation technique to automatically determine spectroscopic redshifts for LAEs while marginalizing over sky residuals. We use visually inspected spectra of LAEs obtained using the Dark Energy Spectroscopic Instrument (DESI) to create a data-driven prior and can determine redshift by jointly inferring sky residual, LAE, and residual components for each individual spectrum. We demonstrate this method on 881 spectroscopically observed z = 2 − 4 z=2-4 DESI LAE candidate spectra and determine their redshifts with > > 90% accuracy when validated against visually inspected redshifts. Using the Δ ​ χ 2 \Delta\chi^{2} value from our pipeline as a proxy for detection confidence, we then explore potential survey design choices and implications for targeting LAEs with medium-band photometry. This method allows for scalability and accuracy in determining redshifts from DESI spectra, and the results provide recommendations for LAE targeting in anticipation of future high-redshift spectroscopic surveys. Astropy [ 5 ] , Julia [ 7 ] , FastSpecFit [ 56 ] I Introduction Lyman- \alpha emitters (LAEs) are young, star-forming galaxies with significant emission (EW >20 Å) in the Ly \alpha line [ 36 , 58 ] . Their existence was first theorized by [ 63 ] , and since then, modern surveys have inferred the physical properties of LAEs. They are low-metallicity with little dust extinction [ 1 , 59 ] , and have significantly higher star formation rates (SFRs) than expected for galaxies of their size [ 67 ] . They also have relatively low stellar masses [ 32 ] and are thought to be progenitors of present-day Milky Way-like galaxies [ 33 , 58 ] . The Ly \alpha line profile provides clues about underlying physical processes in LAEs [ 48 ] . Studies using theoretical radiative transfer simulations suggest that the line shape correlates with intrinsic properties of the emission, such as Ly \alpha continuum leakage and hydrogen column density [ 86 , 85 ] . There is a known relationship between the FWHM of the Ly \alpha line and a systemic velocity offset, possibly due to the opacity of neutral hydrogen in the LAE’s interstellar medium [ 87 ] . Population studies of LAEs over a large redshift range provide valuable constraints for models of reionization [ 60 , 90 , 58 ] . LAEs are more common at higher redshifts up to around z\approx 6 before their number density gradually reduces when approaching z=7-8 [ 75 , 58 , 82 ] . Similarly, the Ly \alpha luminosity function can be used to constrain the neutral hydrogen fraction x_{\textsc{Hi}}=n_{\textsc{Hi}}/(n_{\textsc{Hi}}+n_{\textsc{Hii}}) , which climbs from zero at z<6 to x_{\textsc{Hi}}>0.6 at z\approx 8 [ 60 , 43 ] . LAEs can be used as spatial tracers of the early universe [ 42 ] . They have a similar bias to low-redshift massive galaxies and thus extend our probes of large-scale structure to higher redshifts [ 88 ] . LAEs in “blobs” closely trace galactic protoclusters [ 26 , 41 , 69 ] . Studies of LAE clustering also provide constraints on the masses of their dark matter halos [ 61 , 37 ] . Large-scale observational studies of LAEs have been key to realizing their potential as cosmological probes. LAEs have been targeted mainly through spectroscopy and narrow-band photometry, and have been observed from the ground and space, including with Keck [ 79 , 27 , 38 , 39 , 81 , 53 , 11 ] , Very Large Telescope [ 46 ] , Hubble Space Telescope [ 64 ] , Dark Energy Camera [ 51 , DECam;] , Subaru Hyper-Suprime Cam [ 44 , 3 , 45 ] , James Webb Space Telescope [ 47 ] , and most recently, the Dark Energy Spectroscopic Instrument [ 29 , DESI;] . The Dark Energy Spectroscopic Instrument (DESI) has collected spectra of millions of galaxies for precise cosmological measurements [ 13 , 16 , 25 ] . While LAEs are not one of the primary targets, they have been observed by DESI through secondary and tertiary targeting programs, providing spectroscopic follow-up for LAE candidates identified from photometric surveys [ 29 , 68 ] . After DESI completes its initial five-year run, the planned follow-up program (DESI-2) will focus on higher-redshift targets and aims to observe \approx 1.5 million LAEs [ 77 , 70 ] . The current DESI spectroscopic data analysis pipeline [ 6 ] does not automatically identify or determine redshifts for LAEs, which are instead currently done by visual inspection. In anticipation of LAEs becoming a primary target for DESI-2, it is important to have reliable, automated software pipelines to identify and characterize them. This requires the use of principled data methods that maximize interpretability while being computationally efficient. Here, we take a component separation approach to LAE identification and automated redshift determination to provide a robust, scalable, and data-driven alternative to visual inspection. Our automated method determines a redshift for each target and provides a \Delta\chi^{2} metric that serves as a proxy for confidence in an LAE detection. We discuss the accuracy of redshift determination and the redshift success rates resulting from targeting LAE targets selected from medium-band (MB) photometry to inform survey design choices in anticipation of DESI-2. An outline of this paper is as follows: in Section II we describe the spectra being used and how the objects were targeted. In Section III we describe in detail the methods we use, how we generate our priors for each component, and how they are applied to the data. The results of the redshift pipeline on real and simulated data and applications to LAE targeting are presented in Section IV and discussed further in Section V before we conclude in Section VI . II Data II.1 DESI DESI is an optical multi-object fiber spectrograph located on the Nicholas U. Mayall 4-meter Telescope at the Kitt Peak National Observatory in Arizona [ 52 , 14 , 16 ] . It uses robotic fiber positioners that allow it to collect 5,000 high-quality spectra simultaneously [ 15 , 78 , 54 ] . DESI spectra span 3600-9800 Å with R=\lambda/\Delta\lambda between 2000 and 5000, and are binned linearly to 0.8 Å pixels [ 13 , 65 ] . For more information about survey validation and the Early Data Release, see DESI Collaboration et al. [17]

Read the complete text