Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ALMA and JWST Identification of Faint Dusty Star-Forming Galaxies up to z~8

Jorge A. Zavala, Andreas L. Faisst, Manuel Aravena, Caitlin M. Casey, Jeyhan S. Kartaltepe, Felix Martinez, John D. Silverman, Sune Toft, Ezequiel Treister, Hollis B. Akins, Hiddo Algera, Karina Barboza, Andrew J. Battisti, Gabriel Brammer, Jackie Champagne, Nicole E. Drakos, Eiichi Egami, Xiaohui Fan, Maximilien Franco, Yoshinobu Fudamoto, Seiji Fujimoto, Steven Gillman, Ghassem Gozaliasl, Santosh Harish, Xiangyu Jin, Koki Kakiichi, Darshan Kakkad, Anton M. Koekemoer, Ruqiu Lin, Daizhong Liu, Arianna S. Long, Georgios E. Magdis, Sinclaire Manning, Crystal L. Martin, Jed McKinney, Romain Meyer, Giulia Rodighiero, Victoria Salazar, David B. Sanders, Marko Shuntov, Margherita Talia, Takumi S. Tanaka, Feige Wang, Wuji Wang, Stephen M. Wilkins, Jinyi Yang, Min S. Yun

Abstract

We exploit a new sample of around 400 bright dusty galaxies from the ALMA CHAMPS Large Program, together with the rich JWST multi-band data products in the COSMOS field, to explore and validate new selection methods for identifying dusty star-forming galaxies (DSFGs). Here, we present an effective empirical selection criterion based on a newly defined parameter: I_star = log(M_star) x log(SFR). Incorporating the F277W-F444W color as a second parameter further improves the purity of the selection. We then apply this method to the COSMOS2025 catalog to search for fainter dusty galaxy candidates below the ALMA CHAMPS detection limit and, through a stacking technique, identify a population of high-redshift (z=6-8) DSFGs with an average flux density of$S_1.2mm = 0.15uJy and a space density of ~6E-6 Mpc^-3. This faint population seems to have been missed by most of the previous submillimeter/millimeter surveys, and ground- and space-based UV-to-NIR surveys. Finally, we discuss the possibility of an evolutionary connection between the z > 10 UV-bright galaxies recently discovered by JWST, the faint dusty z=6-8 galaxies identified here, and the population of z=3-5 massive quiescent galaxies, potentially linked as progenitor-descendant populations based on their abundance, redshifts, and stellar masses.

ALMA and JWST Identification of Faint Dusty Star-Forming Galaxies up to z~8

Abstract

We exploit a new sample of around 400 bright dusty galaxies from the ALMA CHAMPS Large Program, together with the rich JWST multi-band data products in the COSMOS field, to explore and validate new selection methods for identifying dusty star-forming galaxies (DSFGs). Here, we present an effective empirical selection criterion based on a newly defined parameter: I_star = log(M_star) x log(SFR). Incorporating the F277W-F444W color as a second parameter further improves the purity of the selection. We then apply this method to the COSMOS2025 catalog to search for fainter dusty galaxy candidates below the ALMA CHAMPS detection limit and, through a stacking technique, identify a population of high-redshift (z=6-8) DSFGs with an average flux density of$S_1.2mm = 0.15uJy and a space density of ~6E-6 Mpc^-3. This faint population seems to have been missed by most of the previous submillimeter/millimeter surveys, and ground- and space-based UV-to-NIR surveys. Finally, we discuss the possibility of an evolutionary connection between the z > 10 UV-bright galaxies recently discovered by JWST, the faint dusty z=6-8 galaxies identified here, and the population of z=3-5 massive quiescent galaxies, potentially linked as progenitor-descendant populations based on their abundance, redshifts, and stellar masses.

Paper Structure

This paper contains 16 sections, 2 equations, 7 figures.

Table of Contents

  1. Introduction
  2. Sample selection and observations
  3. The ALMA-CHAMPS Galaxy Sample
  4. Identification of JWST counterparts
  5. New DSFG selection diagrams
  6. Color-color and color-magnitude selection diagrams:
  7. SED-based selection diagrams:
  8. On the search for high-redshift dusty galaxies
  9. Selection of faint DSFG candidates
  10. Stacking analysis
  11. Constraining their volume density and potential evolution
  12. Caveats and limitations of redshifts estimates
  13. Summary
  14. Spectroscopic confirmation of three galaxies
  15. A deeper look into the $I_\star$ parameter
  16. ...and 1 more sections

Figures (7)

  • Figure 1: The distribution of galaxies in the COSMOS2025 catalog on the $m_{\rm F277W}$ vs $m_{\rm 277W}-m_{\rm 444W}$ ( left) and $m_{\rm 444W}-m_{\rm 770W}$ vs $m_{\rm 277W}-m_{\rm 444W}$ ( right) planes is illustrated with the gray-scale (the darker the color, the higher the number of galaxies). The lighter gray tones on the right panel is the result of the significant lower number of galaxies detected with MIRI with respect to the NIRCam-detected objects. Galaxies detected at $>5\sigma$ in the CHAMPS survey are represented by the large circles and color-coded according to their COSMOS2025 photometric redshifts. Massive, $z>2$ quiescent galaxies (Baker2025; A. Long et al in preparation) and LRDs (Akins2024) are shown as purple and red dots, respectively.
  • Figure 2: Left: Analog to Figure \ref{['fig:selection_colors_only']} but for the $m_{\rm 277W}-m_{\rm 444W}$ vs $I_\star$ plane. The red dotted lines delineate the locus preferentially occupied by DSFGs, which is also used in §\ref{['secc:highz_candidates']} to select high-redshift dusty candidates (note that LRDs are excluded from this selection by imposing a size cut). Right: The COSMOS2025 galaxies in the $I_\star$--$z$ plane, with contours indicating the source density in steps of 1 dex. Each grid cell is colored according to the corresponding galaxies' mean $m_{\rm 277W}-m_{\rm 444W}$ color. Galaxies with high $I_\star$ values of $\gtrsim15$ have systematically redder colors than those from the bulk population. Other populations of red $m_{\rm 277W}-m_{\rm 444W}$ sources, but with lower $I_\star$ ---including quiescent galaxies and line emitters at specific redshifts--- are also highlighted.
  • Figure 3: The top-right panel show the stacking of the CHAMPS 1.2 mm data at the position of the eighteen $z\approx6-8$ DSFGs candidates identified in § \ref{['secc:highz_candidates']} (and shown on the left, with their respective COSMOS2025 IDs). The $5\sigma$ detection in the stack implies an average flux density of $S_{\rm1.2mm}=0.16\pm0.03\,\rm mJy$, confirming both the effectiveness of our selection method to identify dusty candidates and the existence of a dust-enshrouded population of galaxies within the epoch of reionization. The bottom-right panel shows the results of a similar stacking procedure but centered at the position of random sources from the COSMOS2025 catalog, which results in a non-detection.
  • Figure 4: Stellar masses as a function of redshift for different population of galaxies, including spectroscopically-confirmed massive quiescents at $z>3$ (gold circles; Carnall2024Glazebrook2024deGraaff2025Ito2025), JWST-selected galaxies at $z>10$ (blue circles; ArrabalHaro2023Bunker2023Carniani2024Castellano2024Naidu2025Zavala2025), and our $z\approx6-8$ DSFG candidates (red and gray circles). The gray circles represent the best-fit parameters obtained when including the submillimeter photometry in the SED fitting (see §\ref{['secc:highz_properties']}). We also include the evolutionary track of the five massive galaxies from Carnall2024 (blue/green shaded areas), which nicely connect the three different populations. This supports a potential evolutionary link between the $z>10$ galaxies recently discovered with JWST, $z\gtrsim6$ DSFGs, and $z\approx3-5$ massive quiescent galaxies, all of them having a comparable comoving volume density (see §\ref{['secc:highz_properties']}).
  • Figure 5: COSMOS-3D spectra of three of our 18 DSFG candidates with detected emission lines. From top to bottom, ID$_{\rm COSMOS2025}$=386433 at $z=7.20$ (first reported in Meyer2025), 662086 at $z=5.85$ (confirmed by Akins2024 using NIRSpec observations), and 469510 at $z=5.04$ assuming the line corresponds to $H\alpha$ (with an artifact at around $4.45\mu\rm m$).
  • ...and 2 more figures