Predicting the resolved CO emission of z = 1−3 star-forming galaxies

¹ Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany
² Indian Institute of Science Education and Research (IISER) Tirupati, Rami Reddy Nagar, Karakambadi Road, Mangalam (P.O.), Tirupati 517 507, India
³ European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748 Garching, Germany
⁴ Department of Astronomy, University of Florida, 309 Bryant Space Science Center, Gainesville, FL 32611, USA
⁵ Cosmic Dawn Center at the Niels Bohr Institute, University of Copenhagen and DTU-Space, Technical University of Denmark, Denmark
⁶ Center for AstroPhysical Surveys, National Center for Supercomputing Applications, University of Illinois Urbana-Champaign, Urbana, IL 61801, USA

^★ Corresponding author: ravishankar@mpia.de

Received: 18 November 2024
Accepted: 14 May 2025

Abstract

Context. Resolved observations of the CO emission from z = 1–3 star-forming galaxies are becoming increasingly common, with new high-resolution surveys on the horizon.

Aims. We aim to inform the interpretation of this resolved CO emission by creating synthetic observations and testing to what extent routinely observed CO transitions can be used to trace H₂ across galaxy disks.

Methods. To this end, we extract z = 1−3 massive star-forming galaxies (on and above the main sequence) from the SIMBA cos-mological simulation and predict their spatially resolved CO(1–0)-to-CO(5–4) emission using the SLICK pipeline, which combines sub-resolution modeling of the cloud population with the DESPOTIC spectral line calculation code.

Results. We find that the CO(1–0)-to-H₂ ratio (α_CO) varies significantly within these galaxy disks – from values of ∼1– 5 M_⊙ (K km s⁻¹ pc²)⁻¹ in the central 1–3 kpc of the most massive galaxies to >100 M_⊙ (K km s⁻¹ pc²)⁻¹ at ∼15 kpc. Thus, the use of a single α_CO to derive the H₂ surface density leads to severe underestimates of the H₂ contribution in its outskirts. As expected, higher-J CO transitions trace molecular gas in the centers at higher densities, whereas CO(1–0) better traces the more diffuse, extended molecular gas. We see significant variations in the CO excitation, with CO(3–2)/CO(1–0) line luminosity ratios of the most massive galaxies at z ∼2 declining from ∼0.9 in the galaxy centers to ∼0.1 in the outskirts. On average, line ratios increase substantially toward higher redshifts and lower galaxy stellar masses.

Conclusions. We predict that tracing molecular gas with CO beyond 3–5 kpc of cosmic noon galaxies will be challenging with current facilities due to the drastic increase in α_CO. On average, the half-light radii of all CO transitions up to CO(5–4) are consistent with each other, but are ∼27% smaller than the radii enclosing half the total H₂ mass. The predicted line ratios for the central few kiloparsecs of massive galaxies reach supra-thermal values in warm (∼30 − 100 K), dense (>100 cm⁻³) gas. The increased fraction of dense gas in galaxy centers and toward higher redshifts gives rise to CO excitation gradients.