Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Do cosmological simulations reproduce the [OIII] 88 $\mu$m line emission and properties of JWST-discovered galaxies at $z \geq 12$?

T0 review · 3 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Post-processed models of two state-of-the-art simulation suites find no galaxy matching GS-z14 or GHZ2 at z≥12; the closest analogue appears roughly 300 million years after GS-z14's observed epoch.

desk verdict The headline result—simulations don't currently contain a z≥12 analogue of GS-z14/GHZ2, with the best FIRE match appearing only at z≈8.7—is credible; the no-AGN assumption for GHZ2 is the main load-bearing weak point, not the volume arguments. read the letter →

arxiv 2504.18006 v1 pith:C5RNCKKW submitted 2025-04-25 astro-ph.GA

Shengqi Yang , Adam Lidz , Hui Li , Gergo Popping , Jorge A. Zavala , Guochao Sun This is my paper · ORCID
classification astro-ph.GA
keywords galaxyformationhigh-redshiftgalaxies[OIII]88μmlineemissioncosmologicalsimulationsJWSTALMAepochofreionizationemission-linemodeling
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tests whether current cosmological galaxy-formation simulations can produce the extreme [OIII] 88 μm line emitters that JWST and ALMA have found beyond z=12. It builds post-processed line-emission models on top of two leading simulation suites and compares the predicted star-formation rates, stellar masses, metallicities, [OIII] luminosities, sizes, and line widths with the galaxies GS-z14 (z=14.2) and GHZ2 (z=12.3). Neither suite contains a galaxy directly comparable to either observed source. The one simulated galaxy that does match GS-z14 in most measured properties only does so at z=8.7, about 300 million years after GS-z14 is seen. The authors argue that the gap likely reflects the small volumes and limited mass resolution of the simulations rather than a failure of standard galaxy-formation physics, and they use the analogue to forecast what follow-up ALMA and MIRI observations should see.

What carries the argument

The comparison is carried by HIILines, a post-processing framework in which each star particle in a simulated galaxy is treated as the ionizing source of a uniform, isolated HII region; the code solves for the equilibrium ionization structure and level populations and returns line luminosities from the local gas density, metallicity, and stellar population. For the larger-volume IllustrisTNG suite, a Galaxy Mixture Density Network trained on FIRE assigns conditional distributions of line-luminosity-to-stellar-mass ratios to TNG star particles, so the FIRE-based model is effectively transferred onto TNG's galaxy population. This machinery converts simulated galaxies into observable quantities — [OIII] 88 μm luminosity, line width, half-light radius, star-formation rate, and metallicity — making the direct comparison with GS-z14 and GHZ2 possible.

What would settle it

A targeted high-resolution zoom-in simulation of a rare overdense region at $z\approx14$ that yields a galaxy with $\mathrm{SFR}\approx25\,M_\odot\,\mathrm{yr}^{-1}$ and $L_{\mathrm{[OIII]},88}\approx2\times10^8\,L_\odot$ would confirm that the shortfall is a volume effect, whereas a null result would indicate the simulations are missing early-galaxy physics.

Watch

Extended reading notes

Core claim

The paper's central finding is that current simulations do not yet contain a galaxy at z≈12–14 with the [OIII] 88 μm luminosity, star-formation rate, stellar mass, metallicity, size, or line width of GS-z14 or GHZ2. The FIRE galaxy z5m12b comes close to GS-z14 in essentially all measured properties, but at z=8.7 rather than z=14.2, making it a later-forming analogue rather than a direct counterpart. The authors interpret this absence as plausibly due to sampling: the FIRE zoom-in volumes are too small to contain rare sources, and IllustrisTNG's mass resolution is too coarse to resolve galaxies as small as the observed ones. They then use the analogue to predict GS-z14's fluxes in the [OIII] 52 μm, optical [OIII], Hα, and Hβ lines and to estimate the exposure times needed for detection.

Load-bearing premise

The comparison assumes that all the ionizing radiation in GS-z14 and GHZ2 comes from stars, with active galactic nuclei contributing negligibly; if an AGN is hidden in either galaxy, the inferred star-formation rates, metallicities, and the match to stellar-driven simulated HII regions no longer follow.

Editorial extensions

If this is right

  • Neither FIRE nor IllustrisTNG currently produces a galaxy whose [OIII] 88 μm luminosity and star-formation rate match GS-z14 or GHZ2 at z≥12; the closest match is FIRE's z5m12b at z=8.7.
  • The shortfall is plausibly a sampling effect: an extrapolated [OIII] luminosity function suggests roughly one GS-z14-like source per cube about 280 cMpc on a side at z=14, comparable to the volume of the JADES survey that found GS-z14.
  • Bursty star formation creates a large scatter in [OIII]-to-UV luminosity at fixed UV magnitude, so UV-bright galaxies can be [OIII]-faint and vice versa; this scatter explains some non-detections and sets the detection probability as a function of absolute UV magnitude.
  • For GS-z14, JWST/MIRI should detect the rest-frame optical [OIII] 4960,5007 Å lines in under 10 hours and Hα in about 20 hours, while the [OIII] 52 μm line is not observable from the ground at z=14.2.
  • Detecting those lines would pin down the gas temperature, ionizing-photon production rate, and metallicity of GS-z14, testing whether the z5m12b analogue is truly representative.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the volume explanation is right, a targeted zoom-in simulation built on a rare overdense peak at z≈14 should produce a GS-z14 analogue; a null result would point toward missing physics, such as feedback-free starbursts or incomplete treatment of the earliest star formation.
  • The paper's size discrepancy — simulated half-light radii exceed the observed values by factors of 1.3–60, and no viewing direction reproduces GHZ2's 100 pc size — may indicate that z≈14 analogues are intrinsically more compact than later systems, or that the observations are missing low-surface-brightness outskirts; deeper imaging could distinguish these.
  • The analogue predicts an optical [OIII]/Hβ ratio of about 7, which conflicts with the SED-based estimate of Z≈0.02 Z⊙ for GS-z14; a MIRI detection of these lines would directly adjudicate between the two metallicity estimates.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper compares ALMA [OIII] 88 micron detections of GS-z14 at z=14.2 and GHZ2 at z=12.3 against post-processed models from 22 FIRE zoom-in galaxies (using HIILines) and IllustrisTNG galaxies (using the GMDN machine-learning model). It reports that neither simulation suite contains a galaxy at z>=12 that matches the observed SFR, stellar mass, metallicity, [OIII] luminosity, or line width. One FIRE galaxy, z5m12b, matches several of GS-z14's properties at z=8.7, roughly 300 Myr after the observed epoch, though with systematically larger half-light radii. The paper argues that the absence of direct analogues may reflect limited simulation volume and mass resolution rather than incorrect galaxy formation physics, and it uses z5m12b to predict GS-z14's [OIII] 52 micron, rest-optical [OIII], Halpha, and Hbeta fluxes for ALMA and JWST/MIRI follow-up.

Significance. The central negative comparison is a useful, falsifiable constraint: current state-of-the-art simulations do not produce [OIII]-luminous, high-SFR sources like GS-z14 and GHZ2 at z>=12 in the modeled volumes. The paper is strong in using external ALMA and JWST data as fixed benchmarks, with no parameters in the main comparison fitted to either target galaxy, and in making concrete, testable predictions for follow-up observations. The use of open-source HIILines and the public FIRE/IllustrisTNG simulations, together with the careful treatment of viewing-angle distributions for FWHM and R50, is a positive feature. The main weaknesses are the assumed purely stellar ionizing spectrum (especially for GHZ2), the highly extrapolated volume estimate used to explain the absence of analogues, and the unresolved R50 mismatch that complicates the 'closely resembles' claim for z5m12b.

major comments (3)
  1. [Section 2 (and abstract/Section 5)] The comparison rests on the assumption that the ionizing radiation in GS-z14 and GHZ2 is entirely stellar. For GHZ2, the paper itself reports detections of CIV and NIV] and states that the emission 'may partly arise from gas which is photo-ionized by an AGN, or by densely distributed low metallicity stellar populations'; the later dismissal based on extended morphology and low velocity dispersion does not exclude a composite or weak AGN. Since HIILines post-processing uses stellar-population SEDs, a modest AGN contribution (e.g., 10-30% of ionizing photons) would change the predicted [OIII] 88 micron luminosities and the metallicities inferred from line ratios, and the central absence claim for GHZ2 would no longer be a direct test of stellar-driven galaxy formation. Please add a quantitative sensitivity test, for example by recomputing z5m12b's [OIII] luminosity with a power-law ionizing component at varying AGN fractions, and state explicitly how the conclusions change.
  2. [Section 5, volume estimate paragraph] The conclusion that limited volume/mass resolution explains the absence of analogues rests on an estimate that extrapolates the simulated L[OIII]-SFR relation by more than an order of magnitude in SFR, adopts the z~12 UV luminosity function from Sun et al. (2023b) at z=14 with no redshift evolution, and assumes a lognormal scatter of 0.2 dex. The paper labels this a rough estimate, but the final paragraph states that 'it appears plausible' that the mismatch reflects volume effects, which is stronger than the input assumptions warrant. Please present this as one illustrative scenario, vary the assumed UVLF evolution and scatter (e.g., 0.1-0.4 dex), and report the resulting range in the required volume instead of only the best-fit (280 cMpc)^3 and (160 cMpc)^3 values.
  3. [Section 3, Figure 2 and following text] The abstract states that z5m12b 'closely resembles' GS-z14, but Figure 2 shows simulated R50 values exceeding the observed GS-z14 size by factors of 1.3-60 and no viewing direction that reproduces GHZ2's ~100 pc half-light radius. The paper notes the R50 discrepancy and the resulting ~10x dynamical-mass offset, but the 'closely resembles' claim in the abstract is therefore overstated. Please either soften the abstract wording or explicitly quantify the size mismatch in the abstract so that the analogue claim is not misleading.
minor comments (4)
  1. [Various] There are several typographical issues: 'primarilly' in Section 3, 'metallcity' in Section 2, and 'T able' in the Table 1 caption; these should be corrected.
  2. [Figure 2 caption] The caption says 'the colored bands give the measured [OIII] 88 µm line luminosities' while the text describes the red/gold bands as 1-sigma ranges of observed properties; please clarify which bands correspond to which lines and which panels.
  3. [Table 3 and Equation (1)] The sigmoid fits in Table 3 are reported without uncertainties or goodness-of-fit statistics; please provide error bars on a and b and state whether the fits are stable if the assumed 200 km/s line FWHM is varied.
  4. [Figure 1 and Table 1] Table 1 notes that observed luminosities are corrected for gravitational lensing, but Figure 1 does not state whether the simulated luminosities are treated in the same way; please clarify the comparison basis.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: observed ALMA/JWST data are external benchmarks and no parameter in the post-processing is fitted to GS-z14 or GHZ2.

full rationale

The derivation chain is self-contained against external data. The observed [OIII] 88 micron luminosities, SFRs, stellar masses, sizes, FWHMs, and redshifts for GS-z14 and GHZ2 are measured by ALMA and JWST teams independently of the authors' simulation pipeline. HIILines and the GMDN are not recalibrated to GS-z14 or GHZ2 in this paper: HIILines is run on FIRE star particles with parameters set by the simulated gas and stellar properties, and the GMDN was trained on FIRE and applied to IllustrisTNG. The paper cites prior work showing consistency with z~6 ALMA and JWST measurements, which is an external, falsifiable anchor rather than a circular import. The prediction of GS-z14's [OIII] 52 micron and rest-frame optical lines is an extrapolation from the FIRE analogue z5m12b selected by matching GS-z14's SFR, [OIII] 88 micron luminosity, stellar mass, metallicity, and line width; no parameter is fitted to the predicted lines, and the line ratios are computed by HIILines rather than tuned to GS-z14. The Schechter-function volume estimate in Section 5 is explicitly labeled a rough extrapolation, and the conclusion that limited volume and mass resolution may explain the absence is offered as a plausible explanation, not a derived identity. The inferred metallicities use the authors' photoionization method, so the metallicity comparison is method-dependent, but the observed line strengths remain external and the same models are applied consistently to both observations and simulations. The AGN-neglect assumption for GHZ2 is a genuine astrophysical caveat that could affect the comparison, but it is a stated modeling assumption about the ionizing source, not a step that equates output to input by construction. No circular step is present.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central comparison relies on post-processed emission-line models built by the authors (HIILines and GMDN), on an explicit no-AGN assumption, and on transferability of FIRE-trained ML to TNG. No new physical entities are introduced. The volume explanation additionally depends on extrapolated luminosity functions and an assumed lognormal scatter.

free parameters (3)
  • L[OIII]-SFR lognormal scatter = 0.2 dex
    Assumed for the extrapolated [OIII] luminosity function in Section 5; quoted from TNG z=6 results, not directly measured at z=14.
  • Schechter [OIII] luminosity function parameters = phi* = 10^-3.77 cMpc^-3, L* = 10^7.61 Lsun, alpha = -1.70
    Fit to an extrapolated model luminosity function built from FIRE L[OIII]-SFR, the Sun-Furlanetto SFR-UV relation, and the z=12 FIRE UV luminosity function; used to argue that one GS-z14-like source fits in a JADES-like volume.
  • Detection-rate sigmoid parameters (a, b) = Values in Table 3, one pair per redshift from z=6 to z=14
    Fit to GMDN-processed IllustrisTNG galaxies to describe the probability of a 5-sigma ALMA [OIII] detection as a function of MUV; a descriptive fit rather than a physics test.
assumptions (5)
  • domain assumption Stellar radiation is the sole ionizing source in GS-z14 and GHZ2; AGN contributions are neglected.
    Explicitly assumed in Section 2 and used for all metallicity estimates, SFR comparisons, and line predictions. Supported by morphology and velocity dispersion, but not by a full AGN test.
  • domain assumption Each simulated star particle can be treated as an isolated, uniform-density HII region for computing line emission.
    Core approximation in HIILines (Yang et al. 2023); ignores overlapping HII regions, clumping, and multi-phase ISM structure. Previously validated at z=6 and applied here up to z~14 without revalidation.
  • domain assumption A GMDN trained on FIRE zoom-in simulations transfers reliably to IllustrisTNG star particles.
    Needed for the TNG-based detection-rate and luminosity-function statements; the paper notes that TNG cannot resolve bursty star formation and that the model inherits FIRE scatter by construction.
  • domain assumption Underlying FIRE and IllustrisTNG galaxy formation physics, including feedback, cooling, and star formation, is approximately correct at z>=12.
    The conclusion that the gap is mostly volume and resolution rather than a physics failure assumes the simulation physics is not the primary problem.
  • domain assumption Observed half-light radii and FWHMs can be compared directly with simulated 2D projected values over random viewing directions.
    Stated near Figure 2: observational uncertainties in beam deconvolution and spectral resolution are assumed fully accounted for; the size discrepancies may instead reflect observational selection or PSF issues.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Do cosmological simulations reproduce the [OIII] 88 $\mu$m line emission and properties of JWST-discovered galaxies at $z \geq 12$?." pith.science (2026). https://pith.science/paper/C5RNCKKW

@misc{pith2026250418006,
  author       = {Pith},
  title        = {Pith review of: Do cosmological simulations reproduce the [OIII] 88 $\mu$m line emission and properties of JWST-discovered galaxies at $z \geq 12$?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C5RNCKKW}},
  note         = {Machine review of arXiv:2504.18006}
}
abstract

Recent ALMA observations of the [OIII] 88 $\mu$m line provide spectroscopic confirmation of two JWST photometric candidates, GS-z14 and GHZ2, at $z=14.2$ and $z=12.3$, respectively. These discoveries reveal that star formation and chemical enrichment were already underway when the universe was merely 300 Myr old, posing a challenge to galaxy formation models. Here we construct post-processed models for the [OIII] emission lines from galaxies in the state-of-the-art FIRE and IllustrisTNG simulations. Neither simulation suite contains galaxies directly comparable to GS-z14 or GHZ2. However, one simulated FIRE galaxy closely resembles GS-z14 in its star formation rate (SFR), stellar mass, metallicity, [OIII] luminosity and line-width, albeit at $z=8.7$, lagging GS-z14's formation by roughly 300 Myr. Although further investigation is required, we argue that the lack of simulated galaxies matching GS-z14 and GHZ2 may largely be a consequence of the limited volume of the FIRE simulations and the limited mass resolution of Illustris-TNG. We quantify the prospects for follow-up spectroscopic detections of GS-z14 in the [OIII] 52 $\mu$m line with ALMA, and in rest-frame optical [OIII] and Balmer lines with the MIRI instrument on JWST.

Figures

Figures reproduced from arXiv: 2504.18006 by the authors.

Figure 1
Figure 1. A comparison between the observed [OIII] 88µm line luminosity versus SFR relation (Hashimoto et al. 2018; Witstok et al. 2022; Akins et al. 2022; Algera et al. 2024; Fujimoto et al. 2024; Schouws et al. 2024; Zavala et al. 2024b), FIRE simulations (Yang2023), and TNG galaxies (Yang et al. 2024b). The shaded bands at z = 6 give the 1−σ range among TNG50, TNG100, and TNG300 simulated galaxies. At z ≥ 9, the abundance … view at source ↗
Figure 2
Figure 2. The formation and evolution of FIRE galaxy z5m12b. This simulated galaxy broadly resembles GS-z14, except it has similar properties at z = 8.7 rather than at z = 14.2. The panels from top to bottom show (as a function of the age of the universe): the UV luminosity and those in various emission lines; the [OIII] 88 µm line-width (FWHM); the 2D UV and [OIII] half-light radii, R50; the instantaneous SFR; the gas-phase … view at source ↗
Figure 3
Figure 3. Visualizations of z5m12b at z = 13.5 (top) and z = 8.7 (bottom). The LOS direction is selected such that the [OIII] emission half-light radius and FWHM are consis￾tent with GS-z14. In the left panel of each figure the [OIII] surface brightness, the UV continuum surface brightness at 1500˚A, and the gas distribution are shown in the red con￾tours, blue contours, and the gray map, respectively. The right sub-panels sh… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: illustrates the issue using our GMDN￾processed IllustrisTNG galaxies at z = 6, in compari￾son with current measurements.1 The UV luminosities are attenuated according to dust model A from Vogels￾berger et al. (2020). The simulated results are in good agreement with cur…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. The ALPINE-CRISTAL-JWST Survey: The Fast Metal Enrichment of Massive Galaxies at z~5

    astro-ph.GA 2025-10 conditional novelty 6.0 of 10

    Massive galaxies at z~5 were already as enriched in oxygen as their z~2 counterparts, with an FMR whose intrinsic scatter is about five times larger than locally.

Reference graph

Works this paper leans on

47 extracted references · 3 linked inside Pith · cited by 1 Pith paper

  1. [1]

    B., Fujimoto, S., Finlator, K., et al

    Akins, H. B., Fujimoto, S., Finlator, K., et al. 2022, ApJ, 934, 64, doi: 10.3847/1538-4357/ac795b

    doi:10.3847/1538-4357/ac795b 2022

  2. [2]

    Algera, H. S. B., Inami, H., Sommovigo, L., et al. 2024, MNRAS, 527, 6867, doi: 10.1093/mnras/stad3111

    doi:10.1093/mnras/stad3111 2024

  3. [3]

    K., Hashimoto, T., et al

    Binggeli, C., Inoue, A. K., Hashimoto, T., et al. 2021, A&A, 646, A26, doi: 10.1051/0004-6361/202038180 Calabr` o, A., Castellano, M., Zavala, J. A., et al. 2024, ApJ, 975, 245, doi: 10.3847/1538-4357/ad7602

    doi:10.1051/0004-6361/202038180 2021

  4. [4]

    2024a, arXiv e-prints, arXiv:2409.20533, doi: 10.48550/arXiv.2409.20533

    Carniani, S., D’Eugenio, F., Ji, X., et al. 2024a, arXiv e-prints, arXiv:2409.20533, doi: 10.48550/arXiv.2409.20533

    arXiv doi:10.48550/arxiv.2409.20533

  5. [5]

    2024b, Nature, 633, 318, doi: 10.1038/s41586-024-07860-9

    Carniani, S., Hainline, K., D’Eugenio, F., et al. 2024b, Nature, 633, 318, doi: 10.1038/s41586-024-07860-9

    doi:10.1038/s41586-024-07860-9

  6. [6]

    2022, ApJL, 938, L15, doi: 10.3847/2041-8213/ac94d0

    Castellano, M., Fontana, A., Treu, T., et al. 2022, ApJL, 938, L15, doi: 10.3847/2041-8213/ac94d0

    doi:10.3847/2041-8213/ac94d0 2022

  7. [7]

    2024, ApJ, 972, 143, doi: 10.3847/1538-4357/ad5f88

    Castellano, M., Napolitano, L., Fontana, A., et al. 2024, ApJ, 972, 143, doi: 10.3847/1538-4357/ad5f88

    doi:10.3847/1538-4357/ad5f88 2024

  8. [8]

    E., & White, M

    Conroy, C., Gunn, J. E., & White, M. 2009, ApJ, 699, 486, doi: 10.1088/0004-637X/699/1/486

    doi:10.1088/0004-637x/699/1/486 2009

Show all 47 references
  1. [9]

    2023, MNRAS, 523, 3201, doi: 10.1093/mnras/stad1557 D’Eugenio, F., Cameron, A

    Li, Z. 2023, MNRAS, 523, 3201, doi: 10.1093/mnras/stad1557 D’Eugenio, F., Cameron, A. J., Scholtz, J., et al. 2025, ApJS, 277, 4, doi: 10.3847/1538-4365/ada148

    2023 doi

  2. [10]

    J., Johnson, B

    Eisenstein, D. J., Johnson, B. D., Robertson, B., et al. 2023, arXiv e-prints, arXiv:2310.12340, doi: 10.48550/arXiv.2310.12340

    2023 doi

  3. [11]

    S., et al

    Feldmann, R., Boylan-Kolchin, M., Bullock, J. S., et al. 2025, MNRAS, 536, 988, doi: 10.1093/mnras/stae2633

    2025 doi

  4. [12]

    2024, A&A, 689, A310, doi: 10.1051/0004-6361/202450944

    Ferrara, A. 2024, A&A, 689, A310, doi: 10.1051/0004-6361/202450944

    2024 doi

  5. [13]

    2023, MNRAS, 522, 3986, doi: 10.1093/mnras/stad1095

    Ferrara, A., Pallottini, A., & Dayal, P. 2023, MNRAS, 522, 3986, doi: 10.1093/mnras/stad1095

    2023 doi

  6. [14]

    2019, MNRAS, 489, 1, doi: 10.1093/mnras/stz2031

    Ferrara, A., Vallini, L., Pallottini, A., et al. 2019, MNRAS, 489, 1, doi: 10.1093/mnras/stz2031

    2019 doi

  7. [15]

    L., Leung, G

    Finkelstein, S. L., Leung, G. C. K., Bagley, M. B., et al. 2024, ApJL, 969, L2, doi: 10.3847/2041-8213/ad4495

    2024 doi

  8. [16]

    2024, ApJ, 964, 146, doi: 10.3847/1538-4357/ad235c

    Fujimoto, S., Ouchi, M., Nakajima, K., et al. 2024, ApJ, 964, 146, doi: 10.3847/1538-4357/ad235c

    2024 doi

  9. [17]

    2023, ApJS, 265, 5, doi: 10.3847/1538-4365/acaaa9

    Harikane, Y., Ouchi, M., Oguri, M., et al. 2023, ApJS, 265, 5, doi: 10.3847/1538-4365/acaaa9

    2023 doi

  10. [18]

    2018, Nature, 557, 392, doi: 10.1038/s41586-018-0117-z

    Hashimoto, T., Laporte, N., Mawatari, K., et al. 2018, Nature, 557, 392, doi: 10.1038/s41586-018-0117-z

    2018 doi

  11. [19]

    M., Rieke, G

    Helton, J. M., Rieke, G. H., Alberts, S., et al. 2025, Nature Astronomy, doi: 10.1038/s41550-025-02503-z

    2025 doi

  12. [20]

    F., Wetzel, A., Kereˇ s, D., et al

    Hopkins, P. F., Wetzel, A., Kereˇ s, D., et al. 2018, MNRAS, 480, 800, doi: 10.1093/mnras/sty1690

    2018 doi

  13. [21]

    F., Garrison-Kimmel, S., et al

    Ma, X., Hopkins, P. F., Garrison-Kimmel, S., et al. 2018, MNRAS, 478, 1694, doi: 10.1093/mnras/sty1024

    2018 doi

  14. [22]

    C., Casey, C

    Ma, X., Hayward, C. C., Casey, C. M., et al. 2019, MNRAS, 487, 1844, doi: 10.1093/mnras/stz1324

    2019 doi

  15. [23]

    Y., Quataert, E., et al

    Ma, X., Grudi´ c, M. Y., Quataert, E., et al. 2020, MNRAS, 493, 4315, doi: 10.1093/mnras/staa527

    2020 doi

  16. [24]

    C., & Feldmann, R

    Marszewski, A., Sun, G., Faucher-Gigu` ere, C.-A., Hayward, C. C., & Feldmann, R. 2024, ApJL, 967, L41, doi: 10.3847/2041-8213/ad4cee

    2024 doi

  17. [25]

    P., Oesch, P

    Naidu, R. P., Oesch, P. A., van Dokkum, P., et al. 2022, ApJL, 940, L14, doi: 10.3847/2041-8213/ac9b22

    2022 doi

  18. [26]

    D., Tacchella, S., et al

    Robertson, B., Johnson, B. D., Tacchella, S., et al. 2024, ApJ, 970, 31, doi: 10.3847/1538-4357/ad463d

    2024 doi

  19. [27]

    Robertson, B. E. 2022, ARA&A, 60, 121, doi: 10.1146/annurev-astro-120221-044656

    2022 doi

  20. [28]

    L., Shapley, A

    Sanders, R. L., Shapley, A. E., Topping, M. W., Reddy, N. A., & Brammer, G. B. 2024, ApJ, 962, 24, doi: 10.3847/1538-4357/ad15fc

    2024 doi

  21. [29]

    2024, A&A, 682, A98, doi: 10.1051/0004-6361/202346945

    Schimek, A., Decataldo, D., Shen, S., et al. 2024, A&A, 682, A98, doi: 10.1051/0004-6361/202346945

    2024 doi

  22. [30]

    2025, arXiv e-prints, arXiv:2503.10751, doi: 10.48550/arXiv.2503.10751

    Scholtz, J., Parlanti, E., Carniani, S., et al. 2025, arXiv e-prints, arXiv:2503.10751, doi: 10.48550/arXiv.2503.10751

    2025 doi

  23. [31]

    J., Ormerod, K., et al

    Schouws, S., Bouwens, R. J., Ormerod, K., et al. 2024, arXiv e-prints, arXiv:2409.20549, doi: 10.48550/arXiv.2409.20549

    2024 doi arXiv

  24. [32]

    J., Algera, H., et al

    Schouws, S., Bouwens, R. J., Algera, H., et al. 2025, arXiv e-prints, arXiv:2502.01610, doi: 10.48550/arXiv.2502.01610

    2025 doi arXiv

  25. [33]

    2022, A&A, 659, A141, doi: 10.1051/0004-6361/202142786

    Sonnenfeld, A. 2022, A&A, 659, A141, doi: 10.1051/0004-6361/202142786

    2022 doi

  26. [34]

    C., & Shen, X

    Sun, G., Faucher-Gigu` ere, C.-A., Hayward, C. C., & Shen, X. 2023a, MNRAS, 526, 2665, doi: 10.1093/mnras/stad2902 Do cosmological reproduce z ≥ 12 galaxy properties? 11

    doi

  27. [35]

    C., et al

    Sun, G., Faucher-Gigu` ere, C.-A., Hayward, C. C., et al. 2023b, ApJL, 955, L35, doi: 10.3847/2041-8213/acf85a

    doi

  28. [36]

    Sun, G., & Furlanetto, S. R. 2016, MNRAS, 460, 417, doi: 10.1093/mnras/stw980

    2016 doi

  29. [37]

    2019, MNRAS, 484, 5587, doi: 10.1093/mnras/stz243

    Torrey, P., Vogelsberger, M., Marinacci, F., et al. 2019, MNRAS, 484, 5587, doi: 10.1093/mnras/stz243

    2019 doi

  30. [38]

    2020, MNRAS, 492, 5167, doi: 10.1093/mnras/staa137

    Vogelsberger, M., Nelson, D., Pillepich, A., et al. 2020, MNRAS, 492, 5167, doi: 10.1093/mnras/staa137

    2020 doi

  31. [39]

    2022, MNRAS, 515, 1751, doi: 10.1093/mnras/stac1905

    Witstok, J., Smit, R., Maiolino, R., et al. 2022, MNRAS, 515, 1751, doi: 10.1093/mnras/stac1905

    2022 doi

  32. [40]

    2023, ApJL, 942, L9, doi: 10.3847/2041-8213/aca80c

    Yan, H., Ma, Z., Ling, C., Cheng, C., & Huang, J.-S. 2023, ApJL, 942, L9, doi: 10.3847/2041-8213/aca80c

    2023 doi

  33. [41]

    2020, MNRAS, 499, 3417, doi: 10.1093/mnras/staa3000

    Yang, S., & Lidz, A. 2020, MNRAS, 499, 3417, doi: 10.1093/mnras/staa3000

    2020 doi

  34. [42]

    2024a, MNRAS, 534, 3665, doi: 10.1093/mnras/stae2337

    Yang, S., Lidz, A., Benson, A., et al. 2024a, MNRAS, 534, 3665, doi: 10.1093/mnras/stae2337

    doi

  35. [43]

    2023, MNRAS, 525, 5989, doi: 10.1093/mnras/stad2571

    Yang, S., Lidz, A., Smith, A., Benson, A., & Li, H. 2023, MNRAS, 525, 5989, doi: 10.1093/mnras/stad2571

    2023 doi

  36. [44]

    2024b, arXiv e-prints, arXiv:2409.03997, doi: 10.48550/arXiv.2409.03997

    Yang, S., Lidz, A., Benson, A., et al. 2024b, arXiv e-prints, arXiv:2409.03997, doi: 10.48550/arXiv.2409.03997

    doi

  37. [45]

    L., Fujimoto, S., et al

    Yoon, I., Carilli, C. L., Fujimoto, S., et al. 2023, ApJ, 950, 61, doi: 10.3847/1538-4357/acc94d

    2023 doi

  38. [46]

    A., Castellano, M., Akins, H

    Zavala, J. A., Castellano, M., Akins, H. B., et al. 2024a, Nature Astronomy, doi: 10.1038/s41550-024-02397-3

    doi

  39. [47]

    A., Bakx, T., Mitsuhashi, I., et al

    Zavala, J. A., Bakx, T., Mitsuhashi, I., et al. 2024b, ApJL, 977, L9, doi: 10.3847/2041-8213/ad8f38

    doi

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.