Pith. sign in

REVIEW 3 major objections 4 minor 203 references

Simulations with on-the-fly radiation, cosmic rays, and magnetic fields — not boosted supernova feedback — reproduce the UV luminosity function from z≈14 to z≈3 via burstier star formation and ejective-to-preventive outflows.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

T0 review · deepseek-v4-flash

2026-08-01 23:37 UTC pith:Z3WGENQW

load-bearing objection A valuable first large-volume RT+CR+MHD cosmological simulation result whose central causal claim is plausible but not yet isolated from the reduced thermal/kinetic SN budget. the 3 major comments →

arxiv 2607.15344 v1 pith:Z3WGENQW submitted 2026-07-16 astro-ph.GA astro-ph.CO

The Azahar Project: Non-Thermal Physics Drives Star Formation Burstiness and the Evolution of the UV Luminosity Density at Cosmic Dawn

classification astro-ph.GA astro-ph.CO
keywords cosmic raysgalaxy formationhigh-redshift galaxieshydrodynamical simulationsmagnetohydrodynamicsradiative transferreionizationstar formation burstiness
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper argues that cosmological simulations miss the bright, abundant galaxies JWST sees at cosmic dawn because they omit non-thermal feedback channels. It presents the first results of the Azahar suite, zooming in on three models: standard hydrodynamics, hydrodynamics with four-times-boosted supernova energy, and a ‘Non-Thermal’ model that runs radiative transfer, cosmic rays, and magnetic fields on the fly. The Non-Thermal model alone reproduces the observed UV luminosity function from z≈14 to z≈3, the stellar mass function, the star-forming main sequence, and gas metallicities, without tuning any non-thermal parameter. It does so by making star formation burstier and by shifting galaxy outflows from ejective at high redshift to preventive at cosmic noon as cosmic-ray pressure builds up. If the claim holds, JWST’s surprisingly bright early galaxies are not a puzzle for theory but a signature that radiation and cosmic rays regulate the first galaxies.

Core claim

The central claim is that a galaxy formation model with radiative transfer, cosmic-ray hydrodynamics, and magnetic fields running on the fly — the Non-Thermal model — simultaneously matches the observed UV luminosity function from z≈14 to z≈3, the stellar mass function, the star-forming main sequence, and gas-phase metallicities, while standard hydrodynamics overproduces stars and boosted-supernova feedback suppresses early star formation too strongly. The model achieves this through two linked mechanisms: star formation becomes burstier on both long and short timescales, and outflows shift from ejective at high redshift to preventive at cosmic noon as cosmic-ray pressure accumulates in the

What carries the argument

The load-bearing element is the Non-Thermal galaxy formation model: a ~20 pc resolution zoom-in cosmological setup that couples moment-based radiative transfer, a diffusive cosmic-ray fluid, and constrained-transport magnetohydrodynamics to a magneto-thermo-turbulent star formation prescription. Cosmic rays receive 10% of each 10^51 erg supernova and diffuse with a constant coefficient (3×10^28 cm²/s); magnetic fields receive 1%; the rest goes to thermal and kinetic channels. Radiation suppresses star formation on short (~Myr) timescales, while cosmic rays build up a smooth, sustained pressure that supports the ISM and drives temperate outflows. Together they produce bursty, self-regulated g

Load-bearing premise

The paper attributes the success of its Non-Thermal model to radiation, cosmic rays, and magnetic fields, but that run also gives each supernova less energy to thermal and kinetic channels, and the authors have not yet run a control that separates those two changes.

What would settle it

Run a control simulation identical to Non-Thermal but with cosmic-ray and magnetic-field injection disabled while keeping the supernova energy split and all numerical parameters fixed; if that control still matches the UV luminosity function and burstiness from z≈14 to z≈3, the paper’s causal story is wrong.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • If the Non-Thermal result is correct, JWST’s abundant bright galaxies at z>10 arise naturally from bursty, self-regulated star formation rather than requiring new cosmology or extreme feedback-free starbursts.
  • Calibrated boosted-supernova models are disfavored: their outflows remain too ejective at z≈3, over-suppressing star formation and leaving galaxies too metal-poor.
  • Star formation burstiness, measured by the 3-to-50 Myr SFR ratio, becomes a discriminating observable, consistent with current JWST samples and predicted to decrease toward higher stellar mass and lower redshift.
  • The neutral-gas energy budget shifts from thermal to cosmic-ray pressure after z≈7, implying that cosmic rays become a primary ISM support channel in galaxies by cosmic noon.
  • The model predicts a distinctive CGM/IGM signature: smoother Lyα absorption and shallower metal concentration profiles than either standard or calibrated feedback, while preserving filament coherence.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The paper leaves open whether the results are caused by non-thermal physics per se or by the 11% reduction in thermal/kinetic supernova energy; a control run with the same energy split but without cosmic-ray and magnetic injection would settle this.
  • If the ejective-to-preventive transition is real, it should be visible in individual galaxy trajectories — outflow mass loading should decline along merger-free evolutionary tracks, not only in population medians.
  • A testable extension: comparing the B_3/50 burstiness diagnostic with Hα and UV photometry of hundreds of JWST galaxies at z≈6–10 could falsify the timescale structure of the feedback model.
  • The same runs predict LyC photon escape that varies with burst phase, linking the UVLF match to reionization and giving an independent observational probe of non-thermal feedback.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. This paper presents the Azahar suite of high-resolution zoom-in cosmological simulations and compares three models: Standard HD, Calibrated HD (boosted thermal SN feedback), and Non-Thermal (MHD + radiative transfer + cosmic rays on the fly). The central claim is that non-thermal feedback channels produce burstier star formation, drive a transition from ejective to preventive outflows as cosmic-ray pressure builds up, and thereby reproduce the observed UV luminosity function from z~14 to z~3, along with the SFRD, stellar mass function, main sequence, and mass-metallicity relation. The authors argue that simple boosted thermal SN feedback cannot simultaneously match the early and late galaxy populations seen by JWST.

Significance. If the causal claim holds, this is a significant result: a largely uncalibrated model with radiative transfer, cosmic rays, and magnetic fields would naturally explain JWST's bright high-redshift galaxies without artificially boosted thermal feedback. The paper's strengths include a large (~20 pc, ~12 cMpc)^3 sample, a wide set of observational diagnostics, and an explicit, honest treatment of volume and overlap uncertainties. The non-thermal parameters (f_CR, kappa_CR, f_mag) are not fitted to the target observables, which is a genuine methodological strength. However, the central attribution to non-thermal physics is not yet isolated from a simultaneous change in the thermal/kinetic SN energy budget, and the high-redshift UVLF claim is stated more strongly in the abstract than the volume/overlap limitations in the text justify.

major comments (3)
  1. [§2.1.4, Table 1, §3.3, §4] The causal claim is confounded: Non-Thermal changes two things relative to Standard HD—it enables RT/CR/MHD and reduces the thermal+kinetic SN energy budget from 100% to ~89% (10% CR, 1% magnetic). No control run has the same reduced thermal+kinetic budget with all non-thermal channels disabled, and no sensitivity study varies f_CR or kappa_CR. The paper itself states in §3.3 that 'the specific influence of each non-thermal component will be systematically addressed in future work.' As written, the conclusion 'non-thermal physics drives' could instead reflect the energy split or the adopted CR transport parameters. A control run and/or parameter variations, or a substantially softened causal claim, are needed.
  2. [§3.2.1, Fig. 3, footnote 8] The abstract claims the model reproduces the observed UV luminosity function from z~14, but at z>=12.5 the simulated volume primarily samples M_UV ≳ -18, only partially overlapping current JWST measurements. The text explicitly says the faint-end slopes at these redshifts are 'rough estimates' and that the comparison 'focuses on the relative model behavior rather than on the agreement with observations.' The quantitative match at z~14 is therefore not established at the same level as at z~6–3. Please either restrict the high-z claim in the abstract/conclusions or add a JWST-like selection/overlap analysis that makes the comparison direct.
  3. [§2.1.5, §3.4] The non-thermal parameters are fixed to literature values, but the paper's quantitative predictions—especially the cosmic-ray pressure buildup and the ejective-to-preventive outflow transition in §3.4—may be sensitive to choices such as kappa_CR = 3e28 cm^2/s and f_CR = 0.1. No sensitivity runs are shown. Given that these parameters directly control the strength and propagation of cosmic-ray feedback, a variation over the observationally plausible range is needed to demonstrate that the mechanism, rather than the specific parameter values, is responsible for the reported behavior.
minor comments (4)
  1. [Abstract] The abstract refers to the 'full-physics model' while the rest of the paper uses 'Non-Thermal' for the same model. Please make the naming consistent.
  2. [Appendix C] Equation (C2) is labeled C2 but there is no Equation (C1) in the appendix; renumber or add the missing equation.
  3. [§3.3] The text references 'thesan-zoom'; the simulation name should be typeset consistently as 'Thesan-zoom' or similar italic style.
  4. [Fig. 3 caption] Minor typographical spacing issues in the caption (e.g., 'Calibrated HD' appears without spaces in a few places); a proofread pass is recommended.

Circularity Check

0 steps flagged

No significant circularity: the Non-Thermal model's parameters are fixed from independent constraints, and the headline comparisons are external benchmarks rather than fitted targets.

full rationale

The Azahar Non-Thermal simulation is not calibrated to the observed UVLF, stellar mass function, SFRD, main sequence, outflow loading, or metallicities that are used as evidence. Section 2.1.5 states: "none of the non-thermal parameter values described were adjusted to match galaxy property observables; their values are instead motivated by independent observational, theoretical, or numerical constraints," with f_CR = 0.1, kappa_CR = 3e28 cm^2/s, and f_mag = 0.01 tied to external cosmic-ray, supernova-remnant, and magnetic-field constraints. The model is then compared against independent JWST and ground-based datasets (Figs. 3-6, 8, 10, 12); success or failure is not encoded in the model construction. The Calibrated HD model is the only run tuned to a target, and the paper explicitly acknowledges that its z=3 GSMF agreement is "by construction" rather than a prediction. Self-citations (Pandora papers, Dome et al.) appear in motivation and consistency checks, but the central derivation—running an uncalibrated, parameter-fixed model against external benchmarks—does not reduce to those citations. The paper's own admission that "the specific influence of each non-thermal component will be systematically addressed in future work" and the absence of a control run with the same reduced thermal/kinetic SN budget but without RT/CR/MHD weaken the causal attribution to non-thermal physics, but this is a confounded-experiment/correctness concern, not a definitional or fitted-input circularity. No equation or parameter is equivalent to the claimed prediction by construction.

Axiom & Free-Parameter Ledger

6 free parameters · 7 axioms · 0 invented entities

The central claim rests on a large set of sub-grid physical prescriptions and parameter choices. The most important free parameters are the SN energy split (f_CR, f_mag), the CR diffusion coefficient, the seed magnetic field, and the reduced speed of light; none are fitted to the target observables, but no sensitivity analysis is provided. No new physical entities beyond well-established processes (radiation, CRs, magnetic fields) are invoked.

free parameters (6)
  • SN feedback boost α_SN (Calibrated HD) = 4.0
    Calibrated to reproduce the galaxy stellar mass function at z=3; the comparison model's early-Universe behavior is therefore partially constructed from later-time data.
  • CR energy fraction f_CR = 0.1
    Fixed fraction of SN energy injected into cosmic rays; chosen from observational estimates, but is a free input whose value controls CR pressure buildup and the ejective-to-preventive transition.
  • CR diffusion coefficient κ_CR = 3e28 cm^2/s
    Constant isotropic diffusion coefficient from Milky Way gamma-ray constraints; no dependence on plasma environment; directly shapes CR confinement and outflow transition redshift.
  • SN magnetic energy fraction f_mag = 0.01
    SN-injected magnetic energy fraction, chosen to reproduce SN-remnant magnetization; affects magnetic pressure support and star formation efficiency via the MTT prescription.
  • Initial magnetic field B0 = 3e-20 G
    Weak seed field meant to mimic a Biermann battery; its value controls early magnetic amplification timescales in the ISM.
  • Reduced speed of light c~ = 0.01c
    Numerical speed limit for radiative transfer; known to affect ionization front propagation and escape fractions in galaxy simulations.
axioms (7)
  • domain assumption Planck 2016 cosmology and initial conditions generated with MUSIC at z=127.
    The simulation box and zoom region are set up under this cosmology; all derived statistics assume it.
  • domain assumption Star formation follows the magneto-thermo-turbulent (MTT) prescription with local efficiency from turbulent and magnetic pressure.
    The burstiness and star formation rates depend critically on this sub-grid model.
  • domain assumption Mechanical SN feedback with Kroupa IMF, E_SN=1e51 erg, and fixed mass/metal return fractions.
    Supernova feedback is the primary driver of outflows and CR/magnetic injection; the implementation follows Kimm & Cen.
  • domain assumption Radiative transfer solved with a moment method using three hydrogen/helium ionization bins, bpass v2.0 spectra, and reduced speed of light.
    Ionizing radiation is a key early feedback channel; the reduced speed of light and spectral treatment affect when and how radiation regulates star formation.
  • domain assumption Cosmic rays modeled as a single-fluid energy-density component with constant isotropic diffusion and no streaming.
    The claimed CR pressure buildup and ejective-to-preventive transition rely on this transport treatment.
  • domain assumption Haardt & Madau UV background switched on at z=9.
    The UV background affects cooling, heating, and photoionization, especially at later redshifts.
  • domain assumption Effective volume estimated by matching the simulated halo mass function to Sheth-Tormen.
    All number-density statistics (UVLF, GSMF, SFRD) use this volume estimate; its systematic uncertainty is folded into error bands.

pith-pipeline@v1.3.0-alltime-deepseek · 42377 in / 14269 out tokens · 139429 ms · 2026-08-01T23:37:14.308200+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of The Azahar Project: Non-Thermal Physics Drives Star Formation Burstiness and the Evolution of the UV Luminosity Density at Cosmic Dawn." pith.science (2026). https://pith.science/paper/Z3WGENQW

@misc{pith2026260715344,
  author       = {Pith},
  title        = {Pith review of: The Azahar Project: Non-Thermal Physics Drives Star Formation Burstiness and the Evolution of the UV Luminosity Density at Cosmic Dawn},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Z3WGENQW}},
  note         = {Machine review of arXiv:2607.15344}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

JWST observations, which have pushed the discovery and characterization of galaxies to cosmic dawn, have revealed significant deficiencies in state-of-the-art galaxy formation simulations, motivating the need for novel, physically grounded models. We present the first results from the Azahar suite of ten high-resolution ($\sim$20 pc), large-volume zoom-in cosmological simulations, which follow the formation of thousands of galaxies and progressively incorporate radiative transfer (RT), cosmic rays (CRs), and magnetohydrodynamics in addition to the 'standard' baryonic physics. Our Non-Thermal model, which simultaneously includes RT and CRs on the fly, reproduces the observed UV luminosity function from high redshifts ($z \sim 14$) to cosmic noon ($z \sim 3$), as well as the evolution of the stellar mass function, the galaxy main sequence, and observed gas metallicities. It does so through a combination of burstier star formation, with enhanced variability on both long ($\gtrsim 50$ Myr) and short ($\lesssim 10$ Myr) timescales, and a transition from high to low outflow mass-loading factors over cosmic time as CR pressure builds up in the ISM. We find that simple (boosted) thermal SN feedback models fail to capture the cosmic evolution of the very first galaxies because they lack realistic feedback channels that operate on different timescales and have distinct thermodynamical properties, thereby either overproducing stellar mass or driving overly explosive outflows. Our results indicate that non-thermal galaxy formation physics is crucial for providing a robust theoretical framework with which to interpret the high-redshift galaxy populations now being uncovered by JWST.

Figures

Figures reproduced from arXiv: 2607.15344 by Debora Sijacki, Risa H. Wechsler, Sergio Martin-Alvarez.

Figure 1
Figure 1. Figure 1: (Top panel) Projected view showing part of the high-resolution sub-volume (8 cMpc × 3 cMpc × 8 cMpc) for our Non-Thermal model. Colors show the neutral hydrogen (blue), ionized hydrogen (grey), stellar surface density (white), and ionized radiation (orange for low and purple for high energy) densities. Galaxies appear as bright knots across the cosmic web, with the inter-galactic space permeated by Lyman-c… view at source ↗
Figure 2
Figure 2. Figure 2: Comparison of our three illustrative simulations. From top to bottom models are standard hydrodynamics (Stan￾dard HD), calibrated hydrodynamics (Calibrated HD), and Non-Thermal. (Left column) Colors show total gas density (blue/purple), high temperature gas (orange), and stellar surface density (white). Galaxy outflows are very different across models, with Calibrated HD producing much hotter and extended … view at source ↗
Figure 3
Figure 3. Figure 3: UVLF evolution for Standard HD (blue), Calibrated HD (red), and Non-Thermal (green) with recent observational constraints. Redshift decreases from top to bottom, spanning z = 14, 12.5, 11 (left) and z = 10, 6, 3 (right). Shaded bands combine Poisson noise, cosmic variance, and volume uncertainty. Observations are shown as grey open data points. Left panels and top right panel show functional fits from Whit… view at source ↗
Figure 4
Figure 4. Figure 4: Evolution of the cosmic star formation rate den￾sity as inferred from the UV luminosity density. We compare Standard HD (blue), Calibrated HD (red), and Non-Thermal (green) models with various observational estimates (grey data points). Error bands are estimated as for [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Galaxy stellar mass function at z = 10 (top), z = 6 (center), and z = 3 (bottom) for our Standard HD (blue), Calibrated HD (red), and Non-Thermal (green) mod￾els. Error bands are computed as for [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Galaxy main sequence of star-forming galaxies at z = 10 (top left), z = 6 (top right), and z = 3 (bottom) comparing our simulations with observational data (grey points), and empirical predictions (purple). Shaded bands display the 2 σ variance of our measurements across the simulated galaxies. Observational datasets with more than 100 galaxies are summarized with median and quantiles. Thin dotted lines in… view at source ↗
Figure 7
Figure 7. Figure 7: Star formation burstiness (defined as the spectral variance integrated across four characteristic timescale bands; see text) for galaxies with stellar mass M∗(z = 3) > 5×107 M⊙, separated for each of the three sim￾ulations. Error bars show the 1 σ variance across the full dis￾tribution of these galaxies. Timescales are broadly associated with mergers (T ≈ 150 Myr), dynamical cycles (50 Myr), SN feedback (1… view at source ↗
Figure 8
Figure 8. Figure 8: Burstiness parameter B3/50 = SFR3 Myr/ SFR50 Myr (Hα/UV analogue; see Appendix D for the 10/100 Myr defini￾tion, which yields consistent results). We show B3/50 as a function of stellar mass at z = 10 (top left), z = 6 (top right), and z = 3 (bottom) for our standard hydrodynamical (Standard HD, blue), calibrated hydrodynamical (Calibrated HD, red), and full-physics (Non-Thermal, green) models. Lines are s… view at source ↗
Figure 9
Figure 9. Figure 9: Fractional contribution of thermal (fth, red), cosmic ray (fcr, green), and magnetic (fmag, purple) en￾ergy to the energy balance of the H i gas as a function of redshift, for galaxies in the Non-Thermal model with 5×107 ≤ M∗ ≤ 109 M⊙. Lines show the median, and shaded bands the 1 σ scatter across the population. The relative contribution of cosmic rays is secondary at early times and increases toward lowe… view at source ↗
Figure 10
Figure 10. Figure 10: Ionized gas outflow mass-loading factor as a function of stellar mass at z = 10 (top left), z = 6 (top right), and z = 3 (bottom) for the Standard HD (blue), Calibrated HD (red), and Non-Thermal (green) models, compared with observational estimates (grey). Lines are shown as dashed for stellar mass bins with fewer than 10 galaxies, and shaded bands show the 1 σ galaxy population scatter. Thin dotted lines… view at source ↗
Figure 11
Figure 11. Figure 11: Maps of the metal mass surface density (left panels, ΣZ) and Lyα optical depth measured over fixed line-of-sight scales (right panels, τLyα). Panels display the same cosmological region as in [PITH_FULL_IMAGE:figures/full_fig_p020_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: WIM gas mass-metallicity relation at z = 10 (top), z = 6 (center), and z = 3 (bottom), comparing the three simulations (Standard HD in blue, Calibrated HD in red, and Non-Thermal in green) with observational data (grey). Shaded bands correspond to the 1 σ galaxy distri￾bution scatter. Dotted lines in the bottom panel show the evolution of the most massive galaxy in each model. The Non-Thermal model is in … view at source ↗
Figure 13
Figure 13. Figure 13: Comparison of the three volume estimation methods for the Standard HD (blue), Calibrated HD (red), and Non-Thermal (green) simulations. Black lines show the combined estimate for each of the methods: HMF (solid lines), dark matter halos convex hull (dashed lines), and the volume spanned by refined gas cells (dot-dashed lines). The grey shaded band corresponds to the ±1.5σ variance of the combined HMF meas… view at source ↗
Figure 14
Figure 14. Figure 14: Power spectrum of the in-situ SFHs for the three simulations, showing the median variance per logarith￾mic frequency interval, ν Pfrac, as a function of frequency ν, where Pfrac denotes the PSD of the fractional SFH variabil￾ity. The spectrum corresponds to galaxies with stellar mass M∗(z = 3) > 5 × 107 M⊙. each of our models. To compute these spectra, we first compute for each galaxy its precise in-situ … view at source ↗
Figure 15
Figure 15. Figure 15: Comparison of burstiness measurements us￾ing different SFR timescales. Similar to [PITH_FULL_IMAGE:figures/full_fig_p026_15.png] view at source ↗
Figure 17
Figure 17. Figure 17: Lyα optical depth per ∆s = 50 kpc for the Non-Thermal simulation at z = 3, as shown in [PITH_FULL_IMAGE:figures/full_fig_p027_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: Comparison of the mass-metallicity relation shown in [PITH_FULL_IMAGE:figures/full_fig_p027_18.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

203 extracted references · 16 canonical work pages

  1. [1]

    L., & Norman, M

    Abel, T., Bryan, G. L., & Norman, M. L. 2002, Science, 295, 93, doi: 10.1126/SCIENCE.1063991

  2. [2]

    2012, ApJ, 755, 164, doi: 10.1088/0004-637X/755/2/164

    Ackermann, M., Ajello, M., Allafort, A., et al. 2012, ApJ, 755, 164, doi: 10.1088/0004-637X/755/2/164

  3. [3]

    V., Leitner, S

    Agertz, O., Kravtsov, A. V., Leitner, S. N., & Gnedin, N. Y. 2013, ApJ, 770, 25, doi: 10.1088/0004-637X/770/1/25

  4. [4]

    P., Mac Low, M

    Andersson, E. P., Mac Low, M. M., Agertz, O., Renaud, F., & Li, H. 2024, A&A, 681, A28, doi: 10.1051/0004-6361/202347792

  5. [5]

    C., & Jiang, Y.-F

    Armillotta, L., Ostriker, E. C., & Jiang, Y.-F. 2021, ApJ, 922, 11, doi: 10.3847/1538-4357/AC1DB2

  6. [6]

    2023, MNRAS, 527, 11372, doi: 10.1093/MNRAS/STAD3902

    Asada, Y., Sawicki, M., Abraham, R., et al. 2023, MNRAS, 527, 11372, doi: 10.1093/MNRAS/STAD3902

  7. [7]

    2021, MNRAS, 504, 2346, doi: 10.1093/mnras/stab1030

    Attia, O., Teyssier, R., Katz, H., et al. 2021, MNRAS, 504, 2346, doi: 10.1093/mnras/stab1030

  8. [8]

    2004, MNRAS, 352, 376, doi: 10.1111/j.1365-2966.2004.07883.x

    Aubert, D., Pichon, C., & Colombi, S. 2004, MNRAS, 352, 376, doi: 10.1111/j.1365-2966.2004.07883.x

  9. [9]

    H., Hearin, A

    Behroozi, P., Wechsler, R. H., Hearin, A. P., & Conroy, C. 2019, MNRAS, 488, 3143, doi: 10.1093/mnras/stz1182

  10. [10]

    2026, A&A, 707, A396, doi: 10.1051/0004-6361/202557901

    Paladino, R. 2026, A&A, 707, A396, doi: 10.1051/0004-6361/202557901

  11. [11]

    J., Illingworth, G

    Bouwens, R. J., Illingworth, G. D., Oesch, P. A., et al. 2015, ApJ, 803, 34, doi: 10.1088/0004-637X/803/1/34

  12. [12]

    J., Aravena, M., Decarli, R., et al

    Bouwens, R. J., Aravena, M., Decarli, R., et al. 2016, ApJ, 833, 72, doi: 10.3847/1538-4357/833/1/72

  13. [13]

    J., Oesch, P

    Bouwens, R. J., Oesch, P. A., Stefanon, M., et al. 2021, AJ, 162, 47, doi: 10.3847/1538-3881/ABF83E

  14. [14]

    2023, Nature Astronomy, 7, 731, doi: 10.1038/s41550-023-01937-7

    Boylan-Kolchin, M. 2023, Nature Astronomy, 7, 731, doi: 10.1038/s41550-023-01937-7

  15. [15]

    H., et al

    Brauer, K., Mead, J., Wise, J. H., et al. 2025, ApJ, 993, 2, doi: 10.3847/1538-4357/ae06a4

  16. [16]

    L., & Norman, M

    Bryan, G. L., & Norman, M. L. 1998, ApJ, 495, 80, doi: 10.1086/305262

  17. [17]

    S., & Quinn, T

    Butsky, I. S., & Quinn, T. R. 2018, ApJ, 868, 108, doi: 10.3847/1538-4357/aaeac2

  18. [18]

    M., Stanway, E

    Byrne, C. M., Stanway, E. R., Eldridge, J. J., McSwiney, L., & Townsend, O. T. 2022, MNRAS, 512, 5329, doi: 10.1093/MNRAS/STAC807

  19. [19]

    P., et al

    Cadiou, C., Katz, H., Rey, M. P., et al. 2025, arXiv. https://arxiv.org/pdf/2510.05667

  20. [20]

    C., Engelbracht, C

    Calzetti, D., Kennicutt, R. C., Engelbracht, C. W., et al. 2007, ApJ, 666, 870, doi: 10.1086/520082

  21. [21]

    2024, A&A, 685, A99, doi: 10.1051/0004-6361/202347230

    Carniani, S., Venturi, G., Parlanti, E., et al. 2024, A&A, 685, A99, doi: 10.1051/0004-6361/202347230

  22. [22]

    2025, A&A, 704, A290, doi: 10.1051/0004-6361/202556471

    Carvajal-Bohorquez, C., Ciesla, L., Laporte, N., et al. 2025, A&A, 704, A290, doi: 10.1051/0004-6361/202556471

  23. [23]

    2026, MNRAS, 548, doi: 10.1093/MNRAS/STAG740

    Chaikin, E., Schaye, J., Schaller, M., et al. 2026, MNRAS, 548, doi: 10.1093/MNRAS/STAG740

  24. [24]

    2024, ApJL, 976, L15, doi: 10.3847/2041-8213/AD8DC9

    Chemerynska, I., Atek, H., Dayal, P., et al. 2024, ApJL, 976, L15, doi: 10.3847/2041-8213/AD8DC9

  25. [25]

    J., et al

    Chemerynska, I., Atek, H., Furtak, L. J., et al. 2026, MNRAS, 546, doi: 10.1093/MNRAS/STAF2267

  26. [26]

    E., Mead, A

    Chisari, N. E., Mead, A. J., Joudaki, S., et al. 2019, The Open Journal of Astrophysics, 2, 9452, doi: 10.21105/astro.1905.06082

  27. [27]

    2018, MNRAS, 481, 1690, doi: 10.1093/MNRAS/STY2380

    Chisholm, J., Tremonti, C., & Leitherer, C. 2018, MNRAS, 481, 1690, doi: 10.1093/MNRAS/STY2380

  28. [28]

    A., Leitherer, C., & Chen, Y

    Chisholm, J., Tremonti, C. A., Leitherer, C., & Chen, Y. 2017, MNRAS, 469, 4831, doi: 10.1093/mnras/stx1164

  29. [29]

    R., Kereˇ s, D., Hopkins, P

    Choban, C. R., Kereˇ s, D., Hopkins, P. F., et al. 2022, MNRAS, 514, 4506, doi: 10.1093/MNRAS/STAC1542

  30. [30]

    E., Lam, N., et al

    Clarke, L., Shapley, A. E., Lam, N., et al. 2025, arXiv. https://arxiv.org/pdf/2510.06681

  31. [31]

    W., Papovich, C., Finkelstein, S

    Cole, J. W., Papovich, C., Finkelstein, S. L., et al. 2025, ApJ, 979, 193, doi: 10.3847/1538-4357/AD9A6A

  32. [32]

    2022, MNRAS, 513, 2535, doi: 10.1093/mnras/stac1026

    Concas, A., Maiolino, R., Curti, M., et al. 2022, MNRAS, 513, 2535, doi: 10.1093/mnras/stac1026

  33. [33]

    A., Schaye, J., Bower, R

    Crain, R. A., Schaye, J., Bower, R. G., et al. 2015, MNRAS, 450, 1937, doi: 10.1093/mnras/stv725

  34. [34]

    C., Stone, E

    Cummings, A. C., Stone, E. C., Heikkila, B. C., et al. 2016, ApJ, 831, 18, doi: 10.3847/0004-637x/831/1/18

  35. [35]

    2020, MNRAS, 492, 821, doi: 10.1093/mnras/stz3379

    Curti, M., Maiolino, R., Cirasuolo, M., et al. 2020, MNRAS, 492, 821, doi: 10.1093/mnras/stz3379

  36. [36]

    2022, MNRAS, 518, 425, doi: 10.1093/MNRAS/STAC2737

    Curti, M., D’Eugenio, F., Carniani, S., et al. 2022, MNRAS, 518, 425, doi: 10.1093/MNRAS/STAC2737

  37. [37]

    2024, A&A, 684, A75, doi: 10.1051/0004-6361/202346698

    Curti, M., Maiolino, R., Curtis-Lake, E., et al. 2024, A&A, 684, A75, doi: 10.1051/0004-6361/202346698

  38. [38]

    2025, ApJ, 980, 197, doi: 10.3847/1538-4357/ADAB72

    Lopez-Rodriguez, E. 2025, ApJ, 980, 197, doi: 10.3847/1538-4357/ADAB72

  39. [39]

    L., Tacchella, S., ¨Ubler, H., et al

    Danhaive, A. L., Tacchella, S., ¨Ubler, H., et al. 2025, MNRAS, 543, 3249, doi: 10.1093/mnras/staf1540 Bursty First Galaxies in the Azahar Simulations29

  40. [40]

    2020, A&A, 638, A123, doi: 10.1051/0004-6361/201936339 Dav´ e, R., Angl´ es-Alc´ azar, D., Narayanan, D., et al

    Dashyan, G., & Dubois, Y. 2020, A&A, 638, A123, doi: 10.1051/0004-6361/201936339 Dav´ e, R., Angl´ es-Alc´ azar, D., Narayanan, D., et al. 2019, MNRAS, 486, 2827, doi: 10.1093/MNRAS/STZ937

  41. [41]

    2017, A&A, 605, A70, doi: 10.1051/0004-6361/201730419

    Davidzon, I., Ilbert, O., Laigle, C., et al. 2017, A&A, 605, A70, doi: 10.1051/0004-6361/201730419

  42. [42]

    2023, MNRAS, 523, 3201, doi: 10.1093/MNRAS/STAD1557

    Li, Z. 2023, MNRAS, 523, 3201, doi: 10.1093/MNRAS/STAD1557

  43. [43]

    2024, A&A, 691, A231, doi: 10.1051/0004-6361/202450699

    Deng, Y., Li, H., Liu, B., et al. 2024, A&A, 691, A231, doi: 10.1051/0004-6361/202450699

  44. [44]

    2018, ApJS, 239, 35, doi: 10.3847/1538-4365/AAEE8C

    Diemer, B. 2018, ApJS, 239, 35, doi: 10.3847/1538-4365/AAEE8C

  45. [45]

    2025, MNRAS, 537, 629, doi: 10.1093/MNRAS/STAF006

    Sijacki, D. 2025, MNRAS, 537, 629, doi: 10.1093/MNRAS/STAF006

  46. [46]

    T., McLure, R

    Donnan, C. T., McLure, R. J., Dunlop, J. S., et al. 2024, MNRAS, 533, 3222, doi: 10.1093/MNRAS/STAE2037

  47. [47]

    2016, A&A, 585, A138, doi: 10.1051/0004-6361/201527126

    Dubois, Y., & Commer¸ con, B. 2016, A&A, 585, A138, doi: 10.1051/0004-6361/201527126

  48. [48]

    2019, A&A, 631, A121, doi: 10.1051/0004-6361/201936275

    Dubois, Y., Commer¸ con, B., Marcowith, A., & Brahimi, L. 2019, A&A, 631, A121, doi: 10.1051/0004-6361/201936275

  49. [49]

    2014, MNRAS, 444, 1453, doi: 10.1093/mnras/stu1227

    Dubois, Y., Pichon, C., Welker, C., et al. 2014, MNRAS, 444, 1453, doi: 10.1093/mnras/stu1227

  50. [50]

    2021, A&A, 651, A109, doi: 10.1051/0004-6361/202039429

    Dubois, Y., Beckmann, R., Bournaud, F., et al. 2021, A&A, 651, A109, doi: 10.1051/0004-6361/202039429

  51. [51]

    2024, A&A, 687, A240, doi: 10.1051/0004-6361/202449784

    Dubois, Y., Rodr ´ ıguez Montero, F., Guerra, C., et al. 2024, A&A, 687, A240, doi: 10.1051/0004-6361/202449784

  52. [52]

    1992, MNRAS, 256, 43P, doi: 10.1093/mnras/256.1.43P

    Efstathiou, G. 1992, MNRAS, 256, 43P, doi: 10.1093/mnras/256.1.43P

  53. [54]

    2025, ApJ, 987, 189, doi: 10.3847/1538-4357/addc74

    Whitler, L. 2025, ApJ, 987, 189, doi: 10.3847/1538-4357/addc74

  54. [55]

    P., Whitler, L., et al

    Endsley, R., Stark, D. P., Whitler, L., et al. 2023, MNRAS, 524, 2312, doi: 10.1093/mnras/stad1919

  55. [56]

    P., Whitler, L., et al

    Endsley, R., Stark, D. P., Whitler, L., et al. 2024, MNRAS, 533, 1111, doi: 10.1093/MNRAS/STAE1857

  56. [57]

    Bushby, P. J. 2019, MNRAS, 488, 5065, doi: 10.1093/mnras/stz2084

  57. [58]

    Larson, K. L. 2019, ApJ, 884, 133, doi: 10.3847/1538-4357/AB425B

  58. [59]

    2022, MNRAS, 513, 5000, doi: 10.1093/mnras/stac1196

    Martin-Alvarez, S. 2022, MNRAS, 513, 5000, doi: 10.1093/mnras/stac1196

  59. [60]

    2025, A&A, 698, A89, doi: 10.1051/0004-6361/202553924 Faucher-Gigu` ere, C

    Farcy, M., Rosdahl, J., Dubois, Y., et al. 2025, A&A, 698, A89, doi: 10.1051/0004-6361/202553924 Faucher-Gigu` ere, C. A., Quataert, E., & Hopkins, P. F. 2013, MNRAS, 433, 1970, doi: 10.1093/MNRAS/STT866

  60. [61]

    Federrath, C., & Klessen, R. S. 2012, ApJ, 761, 156, doi: 10.1088/0004-637X/761/2/156

  61. [62]

    J., Korista, K

    Ferland, G. J., Korista, K. T., Verner, D. A., et al. 1998, PASP, 110, 761, doi: 10.1086/316190

  62. [63]

    2023, MNRAS, 522, 3986, doi: 10.1093/MNRAS/STAD1095

    Ferrara, A., Pallottini, A., & Dayal, P. 2023, MNRAS, 522, 3986, doi: 10.1093/MNRAS/STAD1095

  63. [64]

    A., Mackey, J., Grassitelli, L., Romano-D ´ ıaz, E., & Porciani, C

    Fichtner, Y. A., Mackey, J., Grassitelli, L., Romano-D ´ ıaz, E., & Porciani, C. 2024, A&A, 690, A72, doi: 10.1051/0004-6361/202449638

  64. [65]

    L., Ryan, R

    Finkelstein, S. L., Ryan, R. E., Papovich, C., et al. 2015, ApJ, 810, 71, doi: 10.1088/0004-637X/810/1/71

  65. [66]

    L., Bagley, M

    Finkelstein, S. L., Bagley, M. B., Ferguson, H. C., et al. 2023, ApJL, 946, L13, doi: 10.3847/2041-8213/acade4

  66. [67]

    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

  67. [68]

    B., Dav´ e, R., Oppenheimer, B

    Ford, A. B., Dav´ e, R., Oppenheimer, B. D., et al. 2014, MNRAS, 444, 1260, doi: 10.1093/mnras/stu1418

  68. [69]

    2006, A&A, 457, 371, doi: 10.1051/0004-6361:20065371

    Fromang, S., Hennebelle, P., & Teyssier, R. 2006, A&A, 457, 371, doi: 10.1051/0004-6361:20065371

  69. [70]

    2015, MNRAS, 454, 4484, doi: 10.1093/mnras/stv2272

    Geen, S., Hennebelle, P., Tremblin, P., & Rosdahl, J. 2015, MNRAS, 454, 4484, doi: 10.1093/mnras/stv2272

  70. [71]

    R., Yan, R., & Tinker, J

    Geha, M., Blanton, M. R., Yan, R., & Tinker, J. L. 2012, ApJ, 757, 85, doi: 10.1088/0004-637X/757/1/85

  71. [72]

    2018, MNRAS, 479, 3042, doi: 10.1093/mnras/sty1653

    Girichidis, P., Naab, T., Hanasz, M., & Walch, S. 2018, MNRAS, 479, 3042, doi: 10.1093/mnras/sty1653

  72. [73]

    2022, MNRAS, 510, 3917, doi: 10.1093/mnras/stab3462

    Girichidis, P., Pfrommer, C., Pakmor, R., & Springel, V. 2022, MNRAS, 510, 3917, doi: 10.1093/mnras/stab3462

  73. [74]

    2010, Nature 2010 463:7278, 463, 203, doi: 10.1038/nature08640

    Governato, F., Brook, C., Mayer, L., et al. 2010, Nature 2010 463:7278, 463, 203, doi: 10.1038/nature08640

  74. [75]

    E., Labbe, I., Goulding, A

    Greene, J. E., Labbe, I., Goulding, A. D., et al. 2024, ApJ, 964, 39, doi: 10.3847/1538-4357/ad1e5f

  75. [77]

    1996, ApJ, 461, 20, doi: 10.1086/177035

    Haardt, F., & Madau, P. 1996, ApJ, 461, 20, doi: 10.1086/177035

  76. [79]

    2013, ApJL, 777, L38, doi: 10.1088/2041-8205/777/2/L38

    Hanasz, M., Lesch, H., Naab, T., et al. 2013, ApJL, 777, L38, doi: 10.1088/2041-8205/777/2/L38

  77. [80]

    2024, ApJ, 960, 56, doi: 10.3847/1538-4357/AD0B7E

    Harikane, Y., Nakajima, K., Ouchi, M., et al. 2024, ApJ, 960, 56, doi: 10.3847/1538-4357/AD0B7E

  78. [81]

    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

  79. [82]

    K., Ellis, R

    Harikane, Y., Inoue, A. K., Ellis, R. S., et al. 2025, ApJ, 980, 138, doi: 10.3847/1538-4357/AD9B2C 30Martin-Alvarez et al

  80. [83]

    J., Adams, N

    Harvey, T., Conselice, C. J., Adams, N. J., et al. 2025, ApJ, 978, 89, doi: 10.3847/1538-4357/AD8C29

Showing first 80 references.