Pith. sign in

REVIEW 2 major objections 6 minor 31 references

After controlling for stellar and halo mass, central black hole mass shows a significant negative partial correlation with cold circumgalactic gas in the IllustrisTNG100 simulation.

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 →

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

T0 review · deepseek-v4-flash

2026-08-01 17:35 UTC pith:4KIV2HOP

load-bearing objection A systematic and useful multiphase partial-correlation census in TNG100; the headline cold-gas signal is likely robust in-simulation, but the SFR>0 exclusion needs a robustness check before the abstract's 'cold gas' claim can be taken fully at face value. the 2 major comments →

arxiv 2607.17567 v1 pith:4KIV2HOP submitted 2026-07-20 astro-ph.GA

The Intrinsic Multiphase Gas--Black Hole Connection across Scales in IllustrisTNG

Xiaoxia Zhang , Taotao Fang , Shulan Yan , Si-Yue Yu , Feng Yuan This is my paper
classification astro-ph.GA
keywords supermassive black holescircumgalactic mediumgalaxy evolutionhydrodynamical simulationsAGN feedbackpartial correlationmultiphase gascold gas
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.

The paper sets out to establish that a supermassive black hole's mass is intrinsically linked to the depletion of cold gas in its host galaxy's circumgalactic medium, independent of stellar mass or dark matter halo mass. Using 5,089 central galaxies from the IllustrisTNG100 simulation at redshift zero, the authors measure the partial correlation between black hole mass and gas mass in four temperature phases (cold, cool, warm, hot) across three radial apertures, after removing the confounding effects of stellar and dark matter mass via group-specific regressions. They find that cold gas within both R200 and 0.15R200 shows a robust negative partial correlation of about -0.37, while warm and hot gas show no substantial intrinsic correlation. The residual pattern indicates that galaxies with overmassive black holes have systematically depleted cold gas, consistent with cumulative AGN feedback. A sympathetic reader would care because this provides a quantitative multiphase diagnostic of AGN feedback and concrete predictions that future multiwavelength surveys can test.

Core claim

After partialling out stellar and dark matter halo mass, the intrinsic correlation between central black hole mass and cold gas mass within 0.15R200 and R200 is significantly negative, with Spearman partial correlation approximately -0.37; cool gas shows a weaker negative signal at R200, warm gas shows no substantial correlation, and hot gas shows only a weak positive correlation at 0.03R200 that vanishes at larger radii. The residual plane for cold gas within R200 reveals a threshold pattern: galaxies whose black holes are overmassive relative to their host stellar and halo mass show systematically depleted cold gas, while under-massive black holes scatter around slightly positive gas resid

What carries the argument

The central tool is a partial correlation analysis: for each galaxy type (star-forming, green valley, quenched), the logarithm of gas mass and the logarithm of black hole mass are separately regressed on the logarithms of stellar mass and dark matter halo mass, and the Spearman rank correlation between the two sets of residuals is the intrinsic gas–black hole connection. The multiphase census is defined by four temperature thresholds (cold T<10^4 K, cool 10^4–10^5 K, warm 10^5–10^6 K, hot ≥10^6 K) and three radial apertures (0.03R200, 0.15R200, R200), with all star-forming gas cells excluded from the gas mass measurements so that temperatures reflect genuine thermal properties.

Load-bearing premise

The cold gas measurement excludes every gas cell with a non-zero star formation rate, and the central claim assumes the removed star-forming gas is not driving the anticorrelation—that the remaining diffuse cold gas is an unbiased tracer of the cold gas reservoir.

What would settle it

Recompute the partial correlation between black hole mass and cold gas mass within R200 including all star-forming gas cells (with their effective equation-of-state temperatures or simply their masses added to the cold phase). If the significant negative partial correlation of about -0.37 weakens to near zero or changes sign, the paper's central claim is refuted as an artifact of the SFR>0 exclusion. A complementary observational check: in a sample of galaxies matched in stellar and halo mass, the predicted negative partial correlation between black hole mass and HI mass should appear; its abs

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

If this is right

  • Black hole mass becomes a predictor of cold circumgalactic gas content on its own, apart from stellar and halo mass, implying the central engine regulates the cold reservoir.
  • The signal strengthens from the inner 0.03R200 aperture out to 0.15R200 and R200, so the feedback imprint is most visible at circumgalactic scales rather than in the inner galactic region.
  • The threshold pattern in the residual plane—overmassive black holes showing systematic cold-gas depletion—implies AGN feedback acts cumulatively rather than as a sharp on/off switch.
  • In quenched galaxies the anticorrelation shifts from cold and cool gas to warm and hot gas, suggesting cumulative heating redistributes the feedback signature across gas phases as star formation ceases.
  • The cold-gas anticorrelation persists across low, medium, and high local density environments with only a weak trend, indicating the AGN signature is largely independent of environment.

Where Pith is reading between the lines

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

  • A robustness test the paper does not perform: including star-forming gas cells in the cold-gas mass could change the result. If the anticorrelation disappears when those cells are included, the claim may be an artifact of excluding dense star-forming gas rather than true depletion of the diffuse cold reservoir.
  • The outward-strengthening cold-gas signal is a specific prediction of the kinetic-mode feedback in this simulation; applying the same partial-correlation framework to simulations with purely thermal AGN feedback should not reproduce this radial pattern, providing a direct model-dependence test.
  • An observable consequence, left implicit: galaxy samples selected for overmassive black holes should show systematically lower neutral hydrogen fractions once stellar and halo mass are matched, a test feasible with existing HI surveys.
  • The null warm-gas result implies that upcoming large absorption-line surveys should stack their samples by black hole mass rather than star formation rate if they hope to detect any warm-gas–black hole connection.

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

2 major / 6 minor

Summary. The manuscript performs a partial-correlation analysis of central black hole mass and multiphase gas content in the IllustrisTNG100 simulation at z=0. Using 5089 central galaxies with M* > 5e9 Msun and MBH > 0, the authors define cold, cool, warm, and hot gas by temperature thresholds, after excluding all gas cells with SFR>0. They measure Spearman partial correlations between log MBH and log Mgas within 0.03R200, 0.15R200, and R200, after controlling for stellar mass and dark matter halo mass via type-specific OLS regressions. The headline result is a significant negative partial correlation (ρ≈−0.37) for cold gas within 0.15R200 and R200, a weaker anti-correlation for cool gas at R200, and null or weak signals for warm and hot gas. The paper further splits the analysis by galaxy type (star-forming, green valley, quenched) and by local environment density, finding that the cold-gas anti-correlation persists across these subsamples, with variation in strength. The authors interpret the result as evidence that AGN feedback progressively depletes cold circumgalactic gas in the TNG model, independent of host stellar and halo mass.

Significance. The paper provides a systematic, quantitative census of the multiphase gas–black hole connection in a widely used public simulation. Its strengths include a transparent and standard partial-correlation methodology, explicit sample-size tables for every phase/aperture bin, bootstrap confidence intervals, and a clear separation of raw and intrinsic correlations. The finding that the cold-gas anti-correlation strengthens outward (from 0.03R200 to R200) and varies across the star-forming–quenched sequence is a useful diagnostic of how TNG's kinetic-mode AGN feedback manifests in the CGM, and it offers a testable baseline for future HI and UV-absorption surveys. However, the central result is largely a restatement of the known design of TNG's feedback model, which was specifically constructed to have massive black holes eject and heat circumgalactic gas; the novelty lies in the multiphase census and the partial-correlation framing rather than in a new physical mechanism. The main concern described below—the exclusion of all star-forming cells from the cold-gas measurement—directly affects whether the headline claim about 'cold gas depletion' is supported.

major comments (2)
  1. [§2.2(ii), §3.2] The cold-gas mass used in the headline result (ρ≈−0.37) excludes every gas cell with SFR>0. In TNG, star-forming cells are exactly the dense, cold ISM, so the measured quantity is not the total cold gas reservoir but its non-star-forming subset. The paper never quantifies the mass fraction removed by this cut, nor whether that fraction is correlated with the MBH residual after controlling for M* and Mdm. If over-massive BHs sit differently along the star-forming–quenched sequence, the split of cold gas into star-forming and non-star-forming cells can differ systematically, so the exclusion alone could generate or amplify the anticorrelation even when the total cold reservoir is not depleted. None of the caveats in §4.3 addresses this selection. A necessary robustness test is to recompute the partial correlations (and residual planes) including star-forming cells, e.g., by using all gas w
  2. [§2.3, Table 1] Table 1 shows that galaxies with zero gas mass in a given phase/aperture bin are omitted from that bin's analysis. For cold gas at 0.03R200, only 2523 of 5089 galaxies enter; for cold gas at R200, 4858 enter. Dropping zeros removes exactly the most gas-poor objects—the population the authors argue is the endpoint of AGN-driven depletion. If the zero/non-zero distinction is informative (e.g., over-massive BHs are more likely to have M_cold=0 after the SFR cut), then the Spearman correlation computed on the truncated sample is a censored estimate and may not reflect the full population. The paper should either include zeros in a rank-based framework, present the zero fraction as a function of BH residual, or demonstrate that the results are robust to the truncation. This is especially relevant for the 0.03R200 scale, where the inferred null result may be an artifact of severe selection.
minor comments (6)
  1. [§4.1] Typo: 'ALF ALF A HIsurvey' should read 'ALFALFA HI survey'.
  2. [References] The Virtanen et al. (2020) reference lists 'Nature Medicine, 17, 261'; the correct journal is Nature Methods.
  3. [Figure 4] The color bar in the green valley heatmap spans −0.4 to 0.4, but the plotted value ρ=−0.41 for cold gas at R200 lies outside this range; this will clip the color encoding and make the value misleading. Extend the color scale or use a diverging normalization that accommodates the full data range.
  4. [§2.2(ii)] The phrase 'ensures that the remaining gas phases reflect genuine thermal properties of the CGM' is only partially accurate: the excluded SF cells include cold dense gas that is part of the ISM/CGM reservoir, and their removal changes the phase budget shown in Fig. 2. Clarify that the phase fractions are for the non-star-forming subset only.
  5. [§3.2 / §2.3] The paper does not apply a multiple-testing correction across the 12 phase-aperture combinations in Fig. 2 or the many entries in Fig. 4. The headline signal is strong enough to survive a Bonferroni correction, but the type-split matrix (Fig. 4) has many tests; a note on this would be helpful.
  6. [§4.3] The third caveat correctly notes that the z=0 snapshot provides only a static endpoint; this is also relevant to the 'threshold pattern' interpretation, which is inferred from a residual plane rather than from time evolution. A sentence acknowledging that the threshold-like pattern could also arise from selection or from the SFR cut would be appropriate.

Circularity Check

0 steps flagged

No circularity: the partial-correlation result is an emergent statistic of TNG, not fixed by the analysis definitions.

full rationale

The paper's derivation chain is an openly specified statistical analysis of IllustrisTNG100 outputs: select central galaxies, define gas phases by temperature after excluding star-forming cells, compute gas masses within three apertures, regress log M_gas and log M_BH on log M_star and log M_dm, then take the Spearman correlation of the residuals. The headline negative partial correlation (rho ≈ -0.37) is not forced by construction: the regressions remove only linear dependence on the two control variables, and the sign and significance of the residual correlation are free outcomes of the simulation data. The cold-gas definition (excluding SFR>0 cells) is a stated scope restriction in Section 2.2(ii), and the paper's claims are explicitly about that non-star-forming cold phase; the absence of a robustness test including star-forming cells is a completeness concern, not a circular reduction. The threshold interpretation is tied to the TNG kinetic-mode feedback model via Weinberger et al. (2017) in Section 4.2, but that is an external model specification rather than a self-citation by the present authors, and measuring the emergent statistical imprint of a feedback model is not circular. No load-bearing self-citations or definitional identifications are present. The analysis is self-contained as a measurement of the simulation, so the appropriate circularity score is 0.

Axiom & Free-Parameter Ledger

0 free parameters · 5 axioms · 0 invented entities

The paper introduces no new entities or fitted constants; it is a statistical measurement on public simulation data. The assumptions listed are the modeling choices that the central claim depends on.

axioms (5)
  • domain assumption TNG's AGN feedback model (Weinberger et al. 2017), especially kinetic-mode winds, is the correct physical representation of BH feedback.
    The entire interpretation of the anticorrelation as AGN feedback relies on this model. Invoked in §2.1 and §4.2.
  • domain assumption Excluding star-forming gas cells yields an unbiased measure of the cold CGM.
    The exclusion removes dense ISM with non-thermodynamic temperatures; no robustness check. Invoked in §2.2(ii).
  • domain assumption Fixed temperature thresholds define physically meaningful phases.
    Phase definitions are conventional, not justified by physical breaks. §2.2(vi).
  • domain assumption Partial correlation controlling for M* and Mdm isolates the intrinsic gas–BH connection.
    Assumes no other major confounders and linear dependence. §2.3.
  • domain assumption The z=0 snapshot is representative of the endpoint of feedback evolution.
    No time evolution is considered; §4.3 third caveat acknowledges this.

pith-pipeline@v1.3.0-alltime-deepseek · 12437 in / 14224 out tokens · 116603 ms · 2026-08-01T17:35:41.840084+00:00 · methodology

0 comments
read the original abstract

The relationship between supermassive black holes and the multiphase circumgalactic medium is central to understanding the co-evolution of galaxies and their central black holes. We investigate this relationship using the IllustrisTNG100 simulation with a sample of 5089 central galaxies at $z=0$, measuring the partial correlation between central black hole mass and the mass of cold ($T < 10^4$K), cool ($10^4 \le T < 10^5$K), warm ($10^5 \le T < 10^6$K), and hot ($T \ge 10^6$K) gas within $0.03R_{200}$, $0.15R_{200}$, and $R_{200}$, after accounting for stellar and dark matter halo mass. We find that after removing these confounding factors, black hole mass shows a significant negative partial correlation ($\rho \approx -0.37$) with cold gas within $R_{200}$ and $0.15R_{200}$, whereas warm and hot gas exhibit no substantial intrinsic correlation. The residual plane reveals a threshold pattern: galaxies with over-massive black holes show systematically reduced cold gas, consistent with the cumulative impact of AGN feedback. The anti-correlation persists across environments with a weak trend in local density, and varies with galaxy type (star-forming, green valley, and quenched). These results provide a quantitative multiphase diagnostic of AGN feedback in TNG and support a picture in which feedback progressively removes cold gas, offering testable predictions for future multiwavelength surveys.

Figures

Figures reproduced from arXiv: 2607.17567 by Feng Yuan, Shulan Yan, Si-Yue Yu, Taotao Fang, Xiaoxia Zhang.

Figure 1
Figure 1. Figure 1: Raw gas–BH correlations at three radial scales. Total gas mass (excluding star-forming gas cells) within 0.03R200 (left), 0.15R200 (middle), and R200 (right) is plotted against black hole mass. Colors indicate galaxy type: star-forming (blue), green valley (green), and quenched (red). Contours (light blue for star-forming, dark red for quenched) show the 2D density distribution; green valley galaxies are n… view at source ↗
Figure 2
Figure 2. Figure 2: Intrinsic partial correlations and gas mass fractions. Left panel: partial correlation heatmap for the full sample, after accounting for log M∗ and log Mdm via group-specific regressions. Colors represent Spearman’s ρ; numbers in cells are ρ values with significance levels (*** for p < 0.001). Right panel: stacked bar chart of gas mass fractions in each temperature phase (Cold: T < 104 K, Cool: 104 ≤ T < 1… view at source ↗
Figure 3
Figure 3. Figure 3: Residual plot for the strongest intrinsic anti– correlation from [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Dependence on star formation activity. Left, middle, and right panels show partial correlation heatmaps for star-forming, green valley, and quenched galaxies, respectively. Significance levels follow the convention of [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Local density effect. Left panel shows the bootstrap distributions of the partial correlation coefficient for cold gas within R200 for low, medium, and high local density groups (based on Σ5), computed using the same group-specific regression and pooling procedure as in [PITH_FULL_IMAGE:figures/full_fig_p007_5.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

31 extracted references · 3 canonical work pages

  1. [1]

    2021, MNRAS, 507, 2383, doi: 10.1093/mnras/stab2310 Astropy Collaboration, Robitaille, T

    Smith, B. 2021, MNRAS, 507, 2383, doi: 10.1093/mnras/stab2310 Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, A33, doi: 10.1051/0004-6361/201322068 Astropy Collaboration, Price-Whelan, A. M., Sip˝ ocz, B. M., et al. 2018, AJ, 156, 123, doi: 10.3847/1538-3881/aabc4f Astropy Collaboration, Price-Whelan, A. M., Lim, P. L., e...

    doi:10.1093/mnras/stab2310 2021

  2. [2]

    Bluck, A. F. L., Piotrowska, J. M., & Maiolino, R. 2023, ApJ, 944, 108, doi: 10.3847/1538-4357/acac7c 10Zhang et al. Bogd´ an,´A., Lovisari, L., Volonteri, M., & Dubois, Y. 2018, ApJ, 852, 131, doi: 10.3847/1538-4357/aa9ab5 Dav´ e, R., Angl´ es-Alc´ azar, D., Narayanan, D., et al. 2019, MNRAS, 486, 2827, doi: 10.1093/mnras/stz937 Dav´ e, R., Crain, R. A.,...

    doi:10.3847/1538-4357/acac7c 2023

  3. [3]

    J., Crain, R

    Davies, J. J., Crain, R. A., Oppenheimer, B. D., & Schaye, J. 2020, MNRAS, 491, 4462, doi: 10.1093/mnras/stz3201 Di Matteo, T., Springel, V., & Hernquist, L. 2005, Nature, 433, 604, doi: 10.1038/nature03335

    doi:10.1093/mnras/stz3201 2020

  4. [4]

    Fabian, A. C. 2012, ARA&A, 50, 455, doi: 10.1146/annurev-astro-081811-125521

    doi:10.1146/annurev-astro-081811-125521 2012

  5. [5]

    L., Werk, J

    Garza, S. L., Werk, J. K., Oppenheimer, B. D., et al. 2024, ApJ, 970, 115, doi: 10.3847/1538-4357/ad4ecc

    doi:10.3847/1538-4357/ad4ecc 2024

  6. [6]

    2019, ApJ, 884, 169, doi: 10.3847/1538-4357/ab3c5d

    Gaspari, M., Eckert, D., Ettori, S., et al. 2019, ApJ, 884, 169, doi: 10.3847/1538-4357/ab3c5d

    doi:10.3847/1538-4357/ab3c5d 2019

  7. [7]

    R., Millman, K

    Harris, C. R., Millman, K. J., van der Walt, S. J., et al. 2020, Nature, 585, 357, doi: 10.1038/s41586-020-2649-2

    doi:10.1038/s41586-020-2649-2 2020

  8. [8]

    Hunter, J. D. 2007, Computing in Science and Engineering, 9, 90, doi: 10.1109/MCSE.2007.55

    doi:10.1109/mcse.2007.55 2007

  9. [9]

    Kormendy, J., & Ho, L. C. 2013, ARA&A, 51, 511, doi: 10.1146/annurev-astro-082708-101811

    doi:10.1146/annurev-astro-082708-101811 2013

  10. [10]

    2019, MNRAS, 488, L134, doi: 10.1093/mnrasl/slz114

    Lakhchaura, K., Truong, N., & Werner, N. 2019, MNRAS, 488, L134, doi: 10.1093/mnrasl/slz114

    work page doi:10.1093/mnrasl/slz114 2019

  11. [11]

    Li, X., Li, C., & Mo, H. J. 2025, ApJ, 980, 145, doi: 10.3847/1538-4357/adace2

    work page doi:10.3847/1538-4357/adace2 2025

  12. [12]

    2018, MNRAS, 480, 5113, doi: 10.1093/mnras/sty2206

    Marinacci, F., Vogelsberger, M., Pakmor, R., et al. 2018, MNRAS, 480, 5113, doi: 10.1093/mnras/sty2206

    doi:10.1093/mnras/sty2206 2018

  13. [13]

    P., Pillepich, A., Springel, V., et al

    Naiman, J. P., Pillepich, A., Springel, V., et al. 2018, MNRAS, 477, 1206, doi: 10.1093/mnras/sty618

    doi:10.1093/mnras/sty618 2018

  14. [14]

    2018, MNRAS, 475, 624, doi: 10.1093/mnras/stx3040

    Nelson, D., Pillepich, A., Springel, V., et al. 2018, MNRAS, 475, 624, doi: 10.1093/mnras/stx3040

    doi:10.1093/mnras/stx3040 2018

  15. [15]

    2019, ApJ, 875, 141, doi: 10.3847/1538-4357/ab107c

    Phipps, F., Bogd´ an,´A., Lovisari, L., et al. 2019, ApJ, 875, 141, doi: 10.3847/1538-4357/ab107c

    doi:10.3847/1538-4357/ab107c 2019

  16. [16]

    2021, MNRAS, 508, 4667, doi: 10.1093/mnras/stab2779

    Pillepich, A., Nelson, D., Truong, N., et al. 2021, MNRAS, 508, 4667, doi: 10.1093/mnras/stab2779

    doi:10.1093/mnras/stab2779 2021

  17. [17]

    2018, MNRAS, 475, 648, doi: 10.1093/mnras/stx3112

    Pillepich, A., Nelson, D., Hernquist, L., et al. 2018, MNRAS, 475, 648, doi: 10.1093/mnras/stx3112

    doi:10.1093/mnras/stx3112 2018

  18. [18]

    M., Bluck, A

    Piotrowska, J. M., Bluck, A. F. L., Maiolino, R., & Peng, Y. 2022, MNRAS, 512, 1052, doi: 10.1093/mnras/stab3673

    doi:10.1093/mnras/stab3673 2022

  19. [19]

    2023, MNRAS, 519, 1526, doi: 10.1093/mnras/stac3214

    Popesso, P., Concas, A., Cresci, G., et al. 2023, MNRAS, 519, 1526, doi: 10.1093/mnras/stac3214

    doi:10.1093/mnras/stac3214 2023

  20. [20]

    2023, MNRAS, 518, 5754, doi: 10.1093/mnras/stac3524

    Ramesh, R., Nelson, D., & Pillepich, A. 2023, MNRAS, 518, 5754, doi: 10.1093/mnras/stac3524

    doi:10.1093/mnras/stac3524 2023

  21. [21]

    A., Bower, R

    Schaye, J., Crain, R. A., Bower, R. G., et al. 2015, MNRAS, 446, 521, doi: 10.1093/mnras/stu2058

    doi:10.1093/mnras/stu2058 2015

  22. [22]

    Silk, J., & Rees, M. J. 1998, A&A, 331, L1, doi: 10.48550/arXiv.astro-ph/9801013

    Pith/arXiv arXiv doi:10.48550/arxiv.astro-ph/9801013 1998

  23. [23]

    2018, MNRAS, 475, 676, doi: 10.1093/mnras/stx3304

    Springel, V., Pakmor, R., Pillepich, A., et al. 2018, MNRAS, 475, 676, doi: 10.1093/mnras/stx3304

    doi:10.1093/mnras/stx3304 2018

  24. [24]

    A., Bell, E

    Terrazas, B. A., Bell, E. F., Pillepich, A., et al. 2020, MNRAS, 493, 1888, doi: 10.1093/mnras/staa374 The pandas development Team. 2025, pandas-dev/pandas: Pandas, v2.3.1, Zenodo, doi: 10.5281/zenodo.3509134

    doi:10.1093/mnras/staa374 2020

  25. [25]

    2021, MNRAS, 501, 2210, doi: 10.1093/mnras/staa3880

    Truong, N., Pillepich, A., & Werner, N. 2021, MNRAS, 501, 2210, doi: 10.1093/mnras/staa3880

    work page doi:10.1093/mnras/staa3880 2021

  26. [26]

    2020, MNRAS, 494, 549, doi: 10.1093/mnras/staa685

    Truong, N., Pillepich, A., Werner, N., et al. 2020, MNRAS, 494, 549, doi: 10.1093/mnras/staa685

    doi:10.1093/mnras/staa685 2020

  27. [27]

    S., & Werk, J

    Tumlinson, J., Peeples, M. S., & Werk, J. K. 2017, ARA&A, 55, 389, doi: 10.1146/annurev-astro-091916-055240

    doi:10.1146/annurev-astro-091916-055240 2017

  28. [28]

    E., et al

    Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature Medicine, 17, 261, doi: 10.1038/s41592-019-0686-2

    doi:10.1038/s41592-019-0686-2 2020

  29. [29]

    2024, Nature, 632, 1009, doi: 10.1038/s41586-024-07821-2

    Wang, T., Xu, K., Wu, Y., et al. 2024, Nature, 632, 1009, doi: 10.1038/s41586-024-07821-2

    doi:10.1038/s41586-024-07821-2 2024

  30. [30]

    2017, MNRAS, 465, 3291, doi: 10.1093/mnras/stw2944

    Weinberger, R., Springel, V., Hernquist, L., et al. 2017, MNRAS, 465, 3291, doi: 10.1093/mnras/stw2944

    doi:10.1093/mnras/stw2944 2017

  31. [31]

    2020, MNRAS, 499, 768, doi: 10.1093/mnras/staa2607

    Zinger, E., Pillepich, A., Nelson, D., et al. 2020, MNRAS, 499, 768, doi: 10.1093/mnras/staa2607

    doi:10.1093/mnras/staa2607 2020