Pith. sign in

REVIEW 2 major objections 5 minor 60 references

Dusty Cloud Acceleration with Multiband Radiation

T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Even a modest ultraviolet component, 5–10 percent of the infrared flux, compresses and disrupts cold dusty clouds so that radiation pressure destroys most dust before the cloud can be significantly accelerated.

desk verdict First systematic multiband UV+IR radiation-hydrodynamics study of dusty cloud acceleration; clean numerics and a plausible sharp UV-disruption threshold, but with an unsecured assumption about dust survival above 1500 K. read the letter →

arxiv 1908.01775 v2 pith:C6ZZ5T6Y submitted 2019-08-05 astro-ph.GA

Xiaoshan Huang , Shane W. Davis , Dong Zhang This is my paper · ORCID
classification astro-ph.GA
keywords galaxies:ISMhydrodynamicsISM:jetsandoutflowsmethod:numericalsimulationradiationpressureondustmultibanddestructioncloudsurvival
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 asks whether radiation pressure on dust can actually accelerate cold gas into galactic outflows, or whether the clouds are destroyed first. Using two-band (ultraviolet and infrared) radiation-hydrodynamic simulations, it finds that ultraviolet light, even at only 5–10 percent of the infrared flux, is disproportionately destructive: it compresses the UV-opaque cloud, drives re-expansion and turbulent mixing with hot background gas, and heats the dust past its destruction temperature before the cloud has moved far. Infrared-only radiation, by contrast, accelerates the cloud more gently and uniformly while re-emitted infrared radiation provides internal pressure support, so the cloud survives much longer. The conclusion is that radiation-pressure driving of cold outflows is most effective where stellar light has already been reprocessed into the infrared, favoring highly obscured galaxies.

What carries the argument

The load-bearing object is the two-band radiation-hydrodynamics setup with a passive dust tracer $s$ that marks cold dusty gas and is set to zero in cells above 1500 K, the assumed dust destruction/decoupling temperature; this tracer defines the cloud mass and the survival time. The dynamical mechanism is differential ultraviolet radiation pressure: with UV opacity $\kappa_{\rm uv}=100\,{\rm cm^2\,g^{-1}}$ versus a low, temperature-dependent Rosseland mean IR opacity, self-shielding makes the illuminated face of the cloud accelerate more than the shielded interior, giving a radiation crushing timescale $t_{\rm rad} = \sqrt{D_c/\Delta a_{\rm uv}}$. The paper uses this timescale, together with the mass-loss curves from the simulations, to argue that compression, re-expansion, and mixing with hot gas—not direct UV heating—destroy the cloud before significant bulk acceleration.

What would settle it

Run the TLMF 10 setup with the dust destruction/decoupling temperature raised to roughly $10^4$ K (or with an explicit grain-sputtering model); if the cloud then retains most of its mass while accelerating past 100 km/s, the 1500 K cutoff is the load-bearing assumption rather than a robust dynamical effect.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that replacing or supplementing infrared radiation with ultraviolet radiation is generally detrimental to dusty cloud survival. Because the cloud is optically thick to UV but optically thin to IR, the UV flux is absorbed near the illuminated surface, creating a radiation-pressure gradient that crushes the cloud; gas pressure then drives re-expansion, and the resulting shear and Rayleigh-Taylor/Kelvin-Helmholtz instabilities mix the outer dusty layers with the hot background, raising them above the assumed 1500 K dust destruction temperature and removing them from the dusty-gas mass. Efficient IR cooling keeps the bulk of the gas near radiative equilibrium (below about 100 K), so destruction is dynamical rather than thermal. Even when UV is only 5–10 percent of the IR flux, this compression–re-expansion–mixing cycle shortens survival to roughly the radiation crushing timescale, whereas IR-only clouds retain most of their mass after several dynamical times and reach comparable velocities. The authors conclude that radiation pressure is most effective when the driving light has been reprocessed into the infrared.

Load-bearing premise

The claim that ultraviolet light destroys clouds before they can be accelerated rests on the assumption that dust is destroyed and decoupled from the gas once mixed gas exceeds about 1500 K.

Editorial extensions

If this is right

  • Radiation-pressure driving of cold molecular outflows will be most efficient in highly obscured galaxies (ULIRGs and high-redshift star formers) where most stellar light is reprocessed into the infrared.
  • In UV-dominated starbursts, cold dusty clouds are shredded within about $10^5$ years and travel only a few parsecs, so they cannot explain the hundreds-of-km/s outflows seen at large radii.
  • Even a UV flux as small as 5–10 percent of the IR flux can cut cloud survival time dramatically, because the UV triggers the compression–re-expansion mixing cycle even when it contributes little to bulk acceleration.
  • IR-only acceleration converts incident momentum into cloud momentum with relatively little disruption, reaching roughly 138 km/s with most of the initial mass intact after $2.5\times10^5$ yr in the fiducial case.
  • Lower UV optical depth makes acceleration faster but disruption also faster, so optically thin UV-driven clouds reach higher velocities yet still lose most of their dusty mass.

Reading between the lines

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

  • A more realistic treatment of grain destruction (e.g., sputtering as grains traverse hot gas) would likely smooth the sharp 1500 K cutoff; because the simulations find that gas crossing 500 K continues heating to above $10^5$ K, the qualitative conclusion should survive, but quantitative survival times could shift.
  • The same compression–re-expansion mechanism should operate in cosmic-ray or hot-wind entrainment contexts whenever an external force accelerates the cloud surface faster than its interior; the survival-time metric used here could be adapted to compare driving mechanisms.
  • Photoionization, which is neglected, would destroy CO and H$_2$ in the UV-illuminated envelope even where dust survives, so the effective molecular-cloud survival time in UV-rich environments is likely shorter than the dusty-gas survival time reported here.
  • The sharp difference between the 1 percent and 5 percent UV-fraction runs suggests an observable diagnostic: galaxies with molecular outflows whose clouds appear filamentary and clumpy may be revealing UV contamination, whereas smoother, longer-lived clouds indicate IR-dominated driving.
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

2 major / 5 minor

Summary. This paper uses two-dimensional and three-dimensional radiation hydrodynamics simulations with Athena++ to study the acceleration of cold, dusty clouds by ultraviolet (UV) and infrared (IR) radiation fields. It introduces a two-band radiation treatment with a passive dust tracer that is destroyed/decoupled above 1500 K, and it analyzes cloud compression, mixing, mass loss, and bulk acceleration. The central finding is that UV radiation, even at about 5-10% of the IR flux, drives a compression/re-expansion cycle that enhances mixing with hot background gas, so that most dusty gas is destroyed before large velocities and flying distances are reached. IR-dominated fields accelerate clouds more gently and preserve them longer. The paper includes analytic estimates for acceleration, equilibrium temperature, and crushing time, and presents resolution, dimensionality, and reduced-speed-of-light checks.

Significance. If the central result holds, it refines previous IR-only radiation-pressure studies by showing that modest UV contamination can reverse the survival advantage of radiation-pressure-driven clouds. This is directly relevant to interpreting molecular outflows in ULIRGs and high-redshift star-forming galaxies, where the UV fraction in launch regions is uncertain. The paper's strengths are its systematic parameter survey, explicit numerical checks in Section 3.4, and parameter-free analytic estimates (Eqs. 17-19) that compare well with simulations. The identified destruction mechanism is physical: radiation-pressure-driven compression and mixing, not direct UV heating, destroys the cloud. The main caveat is that the survival diagnostic is tied to a single assumed dust destruction temperature with one-sided sensitivity tests.

major comments (2)
  1. [Sections 2.1 and 4.1] The dust destruction/decoupling temperature T_dest = 1500 K is load-bearing for the central claim. The passive scalar s is reset to zero in any cell above 1500 K, which both removes that gas from Mc in Eq. (12) and removes its UV/IR opacity from the radiation force via s in Eq. (9). The sensitivity tests described in Section 4.1 vary the threshold downward to 500 K and 1000 K; the argument that gas reaching 500 K quickly exceeds 1500 K explains why lowering the threshold does not matter. It does not, however, test the opposite direction: if dust survives or remains coupled above 1500 K, or if unresolved cold clumps persist in cells whose mean temperature exceeds 1500 K, the surviving mass and flying distance would be larger. Since the abstract and Section 5 conclude that most dust is destroyed before significant acceleration, this assumption should be probed with a higher threshold (e.g., 3000 K) or with a density-dependent decoupling prescription, and the resulting Mc(t), v_mean(t), and flying distance should be reported.
  2. [Section 3.4, Figures 10 and 12] The resolution and dimensionality checks are reassuring at the qualitative level, but the paper's central survival metric is Mc(t). Figure 12 shows that the peak mean density increases monotonically with resolution, and the text acknowledges that the core is not resolved at maximum compression; the 3D run is performed at the resolution of the low-resolution 2D run. Because the rate at which mass is stripped from the core into the mixing layer sets Mc(t), the authors should provide a quantitative comparison of Mc/Mc,0 at selected times (or of the half-mass survival time) for TLUV, TLUV LR, TLUV HR, and TLUV 3D, and state explicitly whether the mass-loss rate converges.
minor comments (5)
  1. [Section 2.1, Eq. (10)] The passive scalar equation has no source term, but the text describes a reset of s to zero above 1500 K; please state that this is applied as an operator-split or post-step reset so the numerical implementation is unambiguous.
  2. [Figure 2 caption] The caption says 'excepting TLUV D and TLUV L has the same a', but those run names do not appear in Table 1; the reference should be to TSUV D and TSUV L.
  3. [Throughout] There are several typographical errors that should be corrected: 'preformed' in Section 3, 'Gallilean' in Section 2.2, 'Kelvin-Helmholz' in Section 3.1, and 'lines of site' in Section 3.3.
  4. [Section 4.1 and Section 3.4] The phrase 'only a modest fraction is a temperatures significantly higher' should read 'at temperatures', and 'follow show little variation' in Section 3.4 should be reworded.
  5. [Equation (18)] The factor of 2 relating E_uv to F_uv/c is introduced without derivation; a one-line explanation would help readers reproduce the Teq estimate.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the UV-disruption finding is an independent simulation outcome, and the 1500 K dust-decoupling reset is a modeling assumption rather than a circular input.

full rationale

The derivation chain is self-contained. The analytic estimates in Eqs. (17)-(19) are parameter-free expressions derived from the adopted opacities, fluxes, and cloud geometry, and the simulations are evolved from stated initial conditions rather than tuned to the conclusion. The central UV-disruption result emerges from simulated compression, re-expansion, and mixing with the hot background, not from the definition of the survival metric; the passive-scalar reset at 1500 K is an assumed dust-destruction/decoupling prescription with external microphysical motivation and one-sided sensitivity tests at 500 K and 1000 K, which is a modeling caveat rather than a reduction of the conclusion to its inputs. Self-citations to Zhang et al. (2018) and related prior work supply the IR-only baseline, the survival-time definition, and some setup choices, but the paper re-runs IR-only cases (TLIR E and TLIR H) independently and the new UV and multiband runs are not forced by those citations. No fitted parameter is renamed as a prediction, and no load-bearing step invokes a self-cited uniqueness theorem. No circular step can be exhibited from the paper's equations or quoted text.

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

The central claim rests on the chosen dust opacity law, the dust destruction temperature, the pressure-equilibrium initial condition, and the reduced-speed-of-light approximation. No new physical entities are introduced. The key free parameters are the UV opacity and the dust destruction temperature, which set the optical depth and the survival metric; both are varied or tested in the paper.

free parameters (5)
  • UV opacity kappa_uv = 100 cm^2/g
    Chosen representative value for dust opacity in the UV, not fitted to data in this paper. It sets the UV optical depth and the UV-to-IR opacity ratio, which the authors note could be explored further (Section 2.1).
  • IR opacity normalization = 10^-3/2 (T/10K)^2 cm^2/g, constant 10^1/2 cm^2/g above 100K
    Rosseland mean opacity approximation from Krumholz & Thompson 2012; a model input rather than something derived in this paper. The normalization affects the IR optical depth and the IR radiation force.
  • Dust destruction/decoupling temperature T_dest = 1500K
    Chosen as the temperature above which the passive dust tracer is set to zero; motivated by grain sublimation but a simplification. The authors test 500K and 1000K and find no significant change in survival times (Section 2.1).
  • Reduced speed of light factor R = 0.01 (0.001 in TLUV R)
    Numerical parameter chosen to accelerate the simulations; validity checked in Section 3.4 by rerunning with R=1e-3. Not a physical parameter, but it is a choice that could affect dynamical evolution if chosen too low.
  • Fiducial incident UV flux F_uv = 1.4 F0 = 4.9e12 L_sun/kpc^2
    Chosen to be favorable to acceleration (about an order of magnitude smaller than typical IR fluxes in ULIRGs). Varying the flux would change the absolute timescales, though the relative UV-vs-IR behavior is the focus.
assumptions (5)
  • domain assumption Initial cloud is in pressure equilibrium with a hot, diffuse background (density contrast 10^5)
    Standard setup in cloud-crushing and entrainment studies (Klein et al. 1994); it sets the environment in which the cloud evolves. Section 2.2.
  • domain assumption Rosseland mean opacity approximation for IR dust opacity (Krumholz & Thompson 2012)
    The opacity law (Eq 9) is taken from Krumholz & Thompson 2012 and Semenov et al. 2003, and is appropriate for T<100K. This is a physical model input, not derived in the paper. Section 2.1.
  • ad hoc to paper Dust is perfectly coupled to gas below 1500K and is destroyed/decoupled above 1500K
    A simple prescription adopted to track cloud survival (Section 2.1). The paper acknowledges more sophisticated grain dynamics would be needed for a definitive treatment.
  • domain assumption Reduced speed of light approximation (R=0.01) preserves the ordering of characteristic timescales
    The validity conditions from Skinner & Ostriker 2013 are assumed to hold; checked by rerunning with R=0.001 (Section 3.4).
  • domain assumption Neglect of self-gravity, magnetic fields, photoionization, conduction, and scattering opacity
    These are simplified away to focus on radiation hydrodynamics; listed as model uncertainties in Section 4.4. The TLUV cloud would be gravitationally unstable if self-gravity were included (Section 3.1), so this is a load-bearing simplification.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Dusty Cloud Acceleration with Multiband Radiation." pith.science (2026). https://pith.science/paper/C6ZZ5T6Y

@misc{pith2026190801775,
  author       = {Pith},
  title        = {Pith review of: Dusty Cloud Acceleration with Multiband Radiation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C6ZZ5T6Y}},
  note         = {Machine review of arXiv:1908.01775}
}
abstract

We perform two-dimensional and three-dimensional simulations of cold, dense clouds, which are accelerated by radiation pressure on dust relative to a hot, diffuse background gas. We examine the relative effectiveness of acceleration by ultraviolet and infrared radiation fields, both independently and acting simultaneously on the same cloud. We study clouds that are optically thin to infrared emission but with varying ultraviolet optical depths. Consistent with previous work, we find relatively efficient acceleration and long cloud survival times when the infrared band flux dominates over the ultraviolet flux. However, when ultraviolet is dominant or even a modest percentage ($\sim 5-10$\%) of the infrared irradiating flux, it can act to compress the cloud, first crushing it and then disrupting the outer layers. This drives mixing of outer regions of the dusty gas with the hot diffuse background to the point where most dust is not likely to survive or stay coupled to the gas. Hence, the cold cloud is unable to survive for a long enough timescale to experience significant acceleration before disruption even though efficient infrared cooling keeps the majority of the gas close to radiative equilibrium temperature ($T \lesssim 100$K). We discuss implications for observed systems, concluding that radiation pressure driving is most effective when the light from star-forming regions is efficiently reprocessed into the infrared.

Figures

Figures reproduced from arXiv: 1908.01775 by the authors.

Figure 1
Figure 1. Simulation snapshots from TLUV. Top panels: density snapshots of both cold and hot gas. Lower panels: temperature of cold dusty gas, the background medium is masked by black. The maximum temperature in color bar corresponds to the temperature at which we set the the passive scalar to zero, representing the overheating of cold gas. t0 = 6.255 × 105 yr, Dc = 0.411pc, ρ0 = 10−19g/cm3 and T0 = 50K. factor of 10 lower th… view at source ↗
Figure 3
Figure 3. Average density (top panel) and temperature (bottom panel) weighted by cold gas density for TLUV (black), TSUV D (green), TSUV L (blue) and TSUV DL (orange). ρinit = ρ0 = 10−19g/cm3 . TSUV D and TSUV DL have lower ρinit = 0.1ρ0 = 10−20g/cm3 . T0 = 50K. The ver￾tical dashed lines in the first row is the radiation crushing time trad for corresponding simulations. Notice that the X axis is scaled to 6.255 × 105 yr. int… view at source ↗
Figure 4
Figure 4. Density snapshots of TSUV L (top panels) ,TSUV D (middle panels) and TSUV DL (bottom panels), the cloud is more optically thin compared to TLUV in these runs. The first row: TSUV L, where the cloud has smaller diameter Dc = 0.1l0. The second row panel: TSUV D is the cloud with lower density ρ = 0.1ρ0. The third row: TSUV DL, the cloud has both lower density and smaller ra￾dius. Notice that the Dc of TSUV L and TSUV … view at source ↗
Figures from the paper (10 more)
Figure 5
Figure 5. Figure 5: Gas pressure distribution of TSUV L (upper left), TSUV DL (upper right), TSUV D (lower left) and TLUV (lower right) at same compression stage. t 0 0 = 0.1t0, D 0 c = 0.1Dc. Characteristic pressure P0 = ρ0v 2 0, P0 = 4.127 × 10−10dyne/cm2 for TSUV L and TLUV, P0 = 4.127…
Figure 6
Figure 6. Figure 6: Mean velocity ∆vmean (top panel), velocity dis￾persion σv (middle panel) and cloud mass Mc (bottom panel) evolution of of TLIR H (red), TLIR E (green) and TLUV (black). v0 = 0.642km/s, and t0 = 6.255 × 105 yr TLUV due to the smaller opacity, which gives rise to an acce…
Figure 9
Figure 9. Figure 9: Average density (top panel) and tempera￾ture (bottom panel) of cold gas for multi-frequency runs TLMF 10 (red), TLMF 5 (blue), TLMF 1 (green) and in￾frared radiation runs TLIR H (black), TLIR E (orange). ρinit = ρ0 = 10−19g/cm3 , T0 = 50K. the TLUV run, the UV radiatio…
Figure 8
Figure 8. Figure 8: Mean velocity ∆vmean (top panel), velocity dispersion σv (middle panel) and cloud mass Mc (bot￾tom panel) evolution of TLMF 10 (red), TLMF 5 (blue), TLMF 1 (green), TLIR H (black) . v0 = 0.642km/s, and t0 = 6.255 × 105 yr TLUV because of the support from re-emitted IR …
Figure 10
Figure 10. Figure 10: Mean velocity ∆vmean (top panel), veloc￾ity dispersion σv (middle panel) and cloud mass Mc (bot￾tom panel) evolution of TLUV 3D (green), TLUV (black) ,TLUV LR (blue), TLUV HR (red), TLUV R (orange). In TLUV R, the radiation flux travels 10 times slower than other runs…
Figure 11
Figure 11. Figure 11: Dust density snapshots of TLUV LR (Left) and TLUV 3D (Right) at t = 0.06t0, the hot background medium is masked by black. t0 = 6.255 × 105 yr, ρ0 = 10−19g/cm3 [PITH_FULL_IMAGE:figures/full_fig_p014_11.png]
Figure 12
Figure 12. Figure 12: Average density (upper panel) and tempera￾ture (lower panel) of cold gas in TLUV 3D (green), TLUV (black) ,TLUV LR (blue), TLUV HR (red), TLUV R (or￾ange). ρinit = ρ0 = 10−19g/cm3 , T0 = 50K. In TLUV R, the radiation flux travels 10 times slower than other runs becaus…
Figure 13
Figure 13. Figure 13: Cold gas density distribution of TLUV at t = 0.04t0, 0.075t0, 0.11t0 (the first, second, third row re￾spectively). The distribution of all dusty material in the calculation domain is the blue solid line (with labels at left). The distribution of hot material, which we…
Figure 14
Figure 14. Figure 14: Cloud mass evolution for different runs. The horizontal grey dashed line labels when cloud mass is half of initial mass, corresponding to the cloud surviving time. Black lines are for TLUV. The blue lines are TSUV L, green lines are TSUV D, and orange lines are TSUV D…
Figure 15
Figure 15. Figure 15: Cloud mass evolution for multi-frequency runs. The horizontal grey dashed line labels half of cloud initial mass, so the time reach it corresponds to the cloud surviving time. The red lines are TLIR H, the black lines are TLUV. The purple lines are the multi-frequency…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 8 canonical work pages

  1. [1]

    2001, ApJL, 556, L11, doi: 10.1086/322860 Banda-Barrag´ an, W

    Weinberg, D. 2001, ApJL, 556, L11, doi: 10.1086/322860 Banda-Barrag´ an, W. E., Parkin, E. R., Federrath, C.,

    doi:10.1086/322860 2001

  2. [2]

    M., & Bicknell, G

    Crocker, R. M., & Bicknell, G. V. 2016, MNRAS, 455, 1309, doi: 10.1093/mnras/stv2405

    doi:10.1093/mnras/stv2405 2016

  3. [3]

    C., & Fabian, A

    Begelman, M. C., & Fabian, A. C. 1990, MNRAS, 244, 26P

    work page 1990

  4. [4]

    J., Bower, R

    Benson, A. J., Bower, R. G., Frenk, C. S., et al. 2003, ApJ, 599, 38, doi: 10.1086/379160

    doi:10.1086/379160 2003

  5. [5]

    D., Warren, S

    Bolatto, A. D., Warren, S. R., Leroy, A. K., et al. 2013, Nature, 499, 450, doi: 10.1038/nature12351

    doi:10.1038/nature12351 2013

  6. [7]

    Bruls, J. H. M. J., Vollm¨ oller, P., & Sch¨ ussler, M. 1999, A&A, 348, 233

    work page 1999

  7. [8]

    A., & Clegg, A

    Chevalier, R. A., & Clegg, A. W. 1985, Nature, 317, 44, doi: 10.1038/317044a0

    doi:10.1038/317044a0 1985

  8. [9]

    2014, A&A, 562, A21, doi: 10.1051/0004-6361/201322464 Dusty Cloud Acceleration 19

    Cicone, C., Maiolino, R., Sturm, E., et al. 2014, A&A, 562, A21, doi: 10.1051/0004-6361/201322464 Dusty Cloud Acceleration 19

    doi:10.1051/0004-6361/201322464 2014

Show all 60 references
  1. [10]

    W., Jiang, Y.-F., Stone, J

    Davis, S. W., Jiang, Y.-F., Stone, J. M., & Murray, N. 2014, ApJ, 796, 107, doi: 10.1088/0004-637X/796/2/107

    2014 doi

  2. [11]

    Draine, B. T. 2011, ApJ, 732, 100, doi: 10.1088/0004-637X/732/2/100

    2011 doi

  3. [12]

    Erb, D. K. 2008, ApJ, 674, 151, doi: 10.1086/524727

    2008 doi

  4. [13]

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

    2012 doi

  5. [14]

    2008, MNRAS, 385, 2181, doi: 10.1111/j.1365-2966.2008.12991.x

    Finlator, K., & Dav´ e, R. 2008, MNRAS, 385, 2181, doi: 10.1111/j.1365-2966.2008.12991.x

    2008

  6. [15]

    E., Tremonti, C., Diamond-Stanic, A

    Geach, J. E., Tremonti, C., Diamond-Stanic, A. M., et al. 2018, ApJL, 864, L1, doi: 10.3847/2041-8213/aad8b6

    2018 doi

  7. [16]

    M., & Best, P

    Heckman, T. M., & Best, P. N. 2014, ARA&A, 52, 589, doi: 10.1146/annurev-astro-081913-035722

    2014 doi

  8. [17]

    2000, ApJS, 129, 493, doi: 10.1086/313421

    Armus, L. 2000, ApJS, 129, 493, doi: 10.1086/313421

    2000 doi

  9. [18]

    F., Quataert, E., & Murray, N

    Hopkins, P. F., Quataert, E., & Murray, N. 2012, MNRAS, 421, 3522, doi: 10.1111/j.1365-2966.2012.20593.x

    2012

  10. [19]

    Ipavich, F. M. 1975, ApJ, 196, 107, doi: 10.1086/153397

    1975 doi

  11. [20]

    M., & Davis, S

    Jiang, Y.-F., Stone, J. M., & Davis, S. W. 2014, ApJS, 213, 7, doi: 10.1088/0067-0049/213/1/7

    2014 doi

  12. [21]

    Kim, J.-G., Kim, W.-T., & Ostriker, E. C. 2016, ApJ, 819, 137, doi: 10.3847/0004-637X/819/2/137 —. 2018, ApJ, 859, 68, doi: 10.3847/1538-4357/aabe27

    2016 doi

  13. [22]

    I., McKee, C

    Klein, R. I., McKee, C. F., & Colella, P. 1994, ApJ, 420, 213, doi: 10.1086/173554

    1994 doi

  14. [23]

    D., Walter, F., et al

    Krieger, N., Bolatto, A. D., Walter, F., et al. 2019, arXiv e-prints, arXiv:1907.00731. https://arxiv.org/abs/1907.00731

    2019 arXiv

  15. [24]

    R., & Thompson, T

    Krumholz, M. R., & Thompson, T. A. 2012, ApJ, 760, 155, doi: 10.1088/0004-637X/760/2/155 —. 2013, MNRAS, 434, 2329, doi: 10.1093/mnras/stt1174

    2012 doi

  16. [25]

    K., Walter, F., Martini, P., et al

    Leroy, A. K., Walter, F., Martini, P., et al. 2015, ApJ, 814, 83, doi: 10.1088/0004-637X/814/2/83

    2015 doi

  17. [26]

    Martin, C. L. 1998, ApJ, 506, 222, doi: 10.1086/306219 —. 2005, ApJ, 621, 227, doi: 10.1086/427277

    1998 doi

  18. [27]

    2015, MNRAS, 449, 2, doi: 10.1093/mnras/stv355

    Quataert, E. 2015, MNRAS, 449, 2, doi: 10.1093/mnras/stv355

    2015 doi

  19. [28]

    2016, A&A, 593, A30, doi: 10.1051/0004-6361/201628978 Mu˜ noz-Vergara, D., Nagar, N

    Rupke, D. 2016, A&A, 593, A30, doi: 10.1051/0004-6361/201628978 Mu˜ noz-Vergara, D., Nagar, N. M., Ramakrishnan, V., et al. 2019, MNRAS, 487, 3679, doi: 10.1093/mnras/stz1572

    2016 doi

  20. [29]

    Murray, N., M´ enard, B., & Thompson, T. A. 2011, ApJ, 735, 66, doi: 10.1088/0004-637X/735/1/66

    2011 doi

  21. [30]

    Murray, N., Quataert, E., & Thompson, T. A. 2005, ApJ, 618, 569, doi: 10.1086/426067

    2005 doi

  22. [31]

    2014, MNRAS, 443, 2018, doi: 10.1093/mnras/stu1271

    Namekata, D., Umemura, M., & Hasegawa, K. 2014, MNRAS, 443, 2018, doi: 10.1093/mnras/stu1271

    2014 doi

  23. [32]

    B., Morganti, R., et al

    Oosterloo, T., Raymond Oonk, J. B., Morganti, R., et al. 2017, A&A, 608, A38, doi: 10.1051/0004-6361/201731781

    2017 doi

  24. [33]

    D., Dav´ e, R., Kereˇ s, D., et al

    Oppenheimer, B. D., Dav´ e, R., Kereˇ s, D., et al. 2010, MNRAS, 406, 2325, doi: 10.1111/j.1365-2966.2010.16872.x

    2010

  25. [34]

    B., Hollenbach, D., Beckwith, S., et al

    Pollack, J. B., Hollenbach, D., Beckwith, S., et al. 1994, ApJ, 421, 615, doi: 10.1086/173677

    1994 doi

  26. [35]

    2014, ApJ, 780, 51, doi: 10.1088/0004-637X/780/1/51

    Smith, D. 2014, ApJ, 780, 51, doi: 10.1088/0004-637X/780/1/51

    2014 doi

  27. [36]

    2015, ApJ, 804, 137, doi: 10.1088/0004-637X/804/2/137

    Proga, D., & Waters, T. 2015, ApJ, 804, 137, doi: 10.1088/0004-637X/804/2/137

    2015 doi

  28. [37]

    2013, MNRAS, 428, 2966, doi: 10.1093/mnras/sts243

    Puchwein, E., & Springel, V. 2013, MNRAS, 428, 2966, doi: 10.1093/mnras/sts243

    2013 doi

  29. [38]

    C., & Skinner, M

    Raskutti, S., Ostriker, E. C., & Skinner, M. A. 2016, ApJ, 829, 130, doi: 10.3847/0004-637X/829/2/130

    2016 doi

  30. [39]

    L., Krumholz, M

    Rosen, A. L., Krumholz, M. R., McKee, C. F., & Klein, R. I. 2016, MNRAS, 463, 2553, doi: 10.1093/mnras/stw2153

    2016 doi

  31. [40]

    S., Veilleux, S., & Sanders, D

    Rupke, D. S., Veilleux, S., & Sanders, D. B. 2002, ApJ, 570, 588, doi: 10.1086/339789

    2002 doi

  32. [41]

    K., & Zweibel, E

    Ruszkowski, M., Yang, H.-Y. K., & Zweibel, E. 2017, ApJ, 834, 208, doi: 10.3847/1538-4357/834/2/208

    2017 doi

  33. [42]

    2015, ApJ, 805, 158, doi: 10.1088/0004-637X/805/2/158

    Scannapieco, E., & Br¨ uggen, M. 2015, ApJ, 805, 158, doi: 10.1088/0004-637X/805/2/158

    2015 doi

  34. [43]

    2018, A&A, 617, A38, doi: 10.1051/0004-6361/201833108

    Schulz, R., Morganti, R., Nyland, K., et al. 2018, A&A, 617, A38, doi: 10.1051/0004-6361/201833108

    2018 doi

  35. [44]

    2003, A&A, 410, 611, doi: 10.1051/0004-6361:20031279

    Sedlmayr, E. 2003, A&A, 410, 611, doi: 10.1051/0004-6361:20031279

    2003 doi

  36. [45]

    L., & Bland-Hawthorn, J

    Shopbell, P. L., & Bland-Hawthorn, J. 1998, ApJ, 493, 129, doi: 10.1086/305108

    1998 doi

  37. [46]

    A., & Ostriker, E

    Skinner, M. A., & Ostriker, E. C. 2013, ApJS, 206, 21, doi: 10.1088/0067-0049/206/2/21 —. 2015, ApJ, 809, 187, doi: 10.1088/0004-637X/809/2/187

    2013 doi

  38. [47]

    W., & Ramirez-Ruiz, E

    Socrates, A., Davis, S. W., & Ramirez-Ruiz, E. 2008, ApJ, 687, 202, doi: 10.1086/590046

    2008 doi

  39. [48]

    S., Aravena, M., B´ ethermin, M., et al

    Spilker, J. S., Aravena, M., B´ ethermin, M., et al. 2018, Science, 361, 1016, doi: 10.1126/science.aap8900

    2018 doi

  40. [49]

    K., & Heckman, T

    Strickland, D. K., & Heckman, T. M. 2007, ApJ, 658, 258, doi: 10.1086/511174 —. 2009, ApJ, 697, 2030, doi: 10.1088/0004-637X/697/2/2030

    2007 doi

  41. [50]

    2019, MNRAS, 488, 1813, doi: 10.1093/mnras/stz1755

    Batcheldor, D. 2019, MNRAS, 488, 1813, doi: 10.1093/mnras/stz1755

    2019 doi

  42. [51]

    A., Quataert, E., Zhang, D., & Weinberg, D

    Thompson, T. A., Quataert, E., Zhang, D., & Weinberg, D. H. 2016, MNRAS, 455, 1830, doi: 10.1093/mnras/stv2428 20

    2016 doi

  43. [52]

    T.-H., & Milosavljevi´ c, M

    Tsang, B. T.-H., & Milosavljevi´ c, M. 2015, MNRAS, 453, 1108, doi: 10.1093/mnras/stv1707

    2015 doi

  44. [53]

    2005, ARA&A, 43, 769, doi: 10.1146/annurev.astro.43.072103.150610

    Veilleux, S., Cecil, G., & Bland-Hawthorn, J. 2005, ARA&A, 43, 769, doi: 10.1146/annurev.astro.43.072103.150610

    2005 arXiv

  45. [54]

    Veilleux, S., Rupke, D. S. N., & Swaters, R. 2009, ApJL, 700, L149, doi: 10.1088/0004-637X/700/2/L149

    2009 doi

  46. [55]

    D., Leroy, A

    Walter, F., Bolatto, A. D., Leroy, A. K., et al. 2017, ApJ, 835, 265, doi: 10.3847/1538-4357/835/2/265

    2017 doi

  47. [56]

    2016, MNRAS, 460, L79, doi: 10.1093/mnrasl/slw056

    Waters, T., & Proga, D. 2016, MNRAS, 460, L79, doi: 10.1093/mnrasl/slw056

    2016 doi

  48. [57]

    G., & Ruszkowski, M

    Wiener, J., Zweibel, E. G., & Ruszkowski, M. 2019, arXiv e-prints, arXiv:1903.01471. https://arxiv.org/abs/1903.01471

    2019 arXiv

  49. [58]

    2018, Galaxies, 6, 114, doi: 10.3390/galaxies6040114

    Zhang, D. 2018, Galaxies, 6, 114, doi: 10.3390/galaxies6040114

    2018 doi

  50. [59]

    Zhang, D., & Davis, S. W. 2017, ApJ, 839, 54, doi: 10.3847/1538-4357/aa6935

    2017 doi

  51. [60]

    W., Jiang, Y.-F., & Stone, J

    Zhang, D., Davis, S. W., Jiang, Y.-F., & Stone, J. M. 2018, ApJ, 854, 110, doi: 10.3847/1538-4357/aaa8e4

    2018 doi

  52. [61]

    A., Murray, N., & Quataert, E

    Zhang, D., Thompson, T. A., Murray, N., & Quataert, E. 2014, ApJ, 784, 93, doi: 10.1088/0004-637X/784/2/93

    2014 doi

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.