Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Characterizing 3D Magnetic Fields and Turbulence in H I Clouds

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A conditional residual neural network can predict the full 3D magnetic field and turbulence properties of diffuse H I clouds from spectroscopic 21-cm observations alone, and the first such maps for two overlapping clouds reveal distinct…

desk verdict A useful DL extension to H I that overclaims the strength of its real-cloud validation; worth refereeing but needs code, uncertainty maps, and an honest reframing. read the letter →

arxiv 2505.07422 v2 pith:GRYSOQTQ submitted 2025-05-12 astro-ph.GA

Yue Hu This is my paper · ORCID
classification astro-ph.GA
keywords interstellarmagneticfieldsHIcloudsMHDturbulencedeeplearning21-cmspectroscopyFASTCRAFTSsurveysonicMachnumberAlfvén
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 sets out to show that the 3D magnetic field of a diffuse atomic hydrogen cloud—its plane-of-sky orientation, line-of-sight inclination, total field strength, sonic Mach number $M_s$, and Alfvén Mach number $M_A$—can be recovered from a single H I spectroscopic observation, using a neural network trained entirely on synthetic data. The authors build a conditional residual neural network, train it on thin velocity channel maps from 27 multiphase 3D MHD simulation setups, and apply it to FAST CRAFTS observations of a low-velocity and an intermediate-velocity H I cloud that overlap on the sky but sit at different distances. They report the first 3D magnetic field characterization for diffuse H I clouds, resolving differences between the two clouds along the line of sight. The result matters because it offers a way to map 3D Galactic magnetic fields from existing 21-cm surveys, with applications to cosmic-ray propagation and CMB foreground removal.

What carries the argument

The central object is a Conditional ResNet, an encoder–decoder convolutional neural network with Feature-wise Linear Modulation (FiLM) conditioning that takes a normalized thin velocity channel map (2 km s$^{-1}$ width) and outputs maps of $\psi$, $\gamma$, $B$, $M_s$, and $M_A$. The physical mechanism carrying the information is the velocity caustic effect: when the channel is narrower than the turbulent velocity dispersion, intensity fluctuations trace MHD eddies that are elongated along the local magnetic field, so the channel morphology encodes field orientation, inclination, and magnetization. The training data come from 54 multiphase MHD simulation snapshots (27 parameter sets covering $B \approx 1, 3, 5$ $\mu$G, three velocity dispersions, and three inclinations), and the model is validated on an unseen simulation with $B \approx 4$ $\mu$G.

What would settle it

Measure the line-of-sight magnetic field toward the low-velocity or intermediate-velocity cloud using Zeeman splitting of H I or Faraday rotation toward background radio sources, and compare with the predicted total field of roughly 5 $\mu$G inclined near 70° to the line of sight; a mismatch substantially larger than the network's reported uncertainties (about 0.2 $\mu$G for $B$ and a few degrees for $\gamma$ on seen data, larger on unseen data) would show the simulation-to-observation transfer fails.

Watch

Extended reading notes

Core claim

The central claim is that the anisotropic imprints of magnetohydrodynamic turbulence in thin H I velocity channels encode enough information to determine the full 3D magnetic field vector and the turbulence state, and that a neural network can decode that information from morphology alone. Trained on synthetic channel maps generated from multiphase AthenaK simulations and applied to FAST data, the network predicts that the low-velocity cloud is nearly trans-Alfvénic and transonic with a field of about 4.8 $\mu$G, while the intermediate-velocity cloud's dense filament is super-Alfvénic ($M_A \approx 1.4$–$1.8$) and supersonic ($M_s \approx 1.5$–$2.0$) with a field of about 5.4 $\mu$G; both clouds sit near 70° inclination to the line of sight. The predicted plane-of-sky angles agree closely with the velocity gradient technique, which independently matches Planck 353 GHz polarization. The paper presents this as the first 3D magnetic field characterization of diffuse H I clouds.

Load-bearing premise

The load-bearing premise is that the synthetic multiphase MHD simulations used to generate the training data faithfully represent the real physical conditions of H I clouds, so that a network trained only on simulated channel maps can predict real magnetic fields without recalibration; only the plane-of-sky angle is checked against independent data.

Editorial extensions

If this is right

  • A single H I datacube can yield the three-dimensional magnetic field geometry and turbulence parameters of diffuse clouds without polarization, Zeeman, or Faraday measurements.
  • Because velocity channels separate gas at different line-of-sight distances, the method can disentangle magnetic fields of clouds that overlap on the sky, as demonstrated for the LVC and IVC.
  • The close agreement of predicted POS angles with Planck polarization suggests the method could help map Galactic polarized foregrounds for CMB studies.
  • Combined with a Galactic rotation curve, the approach could produce 3D magnetic field maps across the Galactic disk, a step the paper explicitly proposes.

Reading between the lines

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

  • I infer the same architecture, retrained on a broader simulation grid, could be applied to all-sky H I surveys to build a 3D magnetic field atlas of the Galaxy; the paper stops at two clouds.
  • The paper validates only the plane-of-sky angle against independent data; a natural extension is to compare predicted field strength and inclination with Zeeman splitting or Faraday rotation measurements toward the same clouds.
  • Its reported super-Alfvénic dense filament in the IVC runs against the usual assumption that dense filaments are strongly magnetized; if confirmed, that would bear on filament formation models, but this consequence is not developed by the author.
  • Because $M_A$ predictions are acknowledged as the least accurate, testing alternative architectures on super-Alfvénic cases would clarify how much of the result is physics versus network capacity.
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

4 major / 5 minor

Summary. The paper presents a Conditional ResNet trained on synthetic H I channel maps generated from 54 multiphase 3D MHD simulation snapshots, spanning a grid of magnetic field strength, velocity dispersion, and inclination angle. The network is designed to predict the plane-of-sky position angle, line-of-sight inclination, magnetic field strength, sonic Mach number, and Alfvén Mach number. The trained model is applied to CRAFTS H I data toward two overlapping Monoceros clouds, an LVC and an IVC, yielding maps of all five quantities. The predicted POS angles are compared with those from the velocity gradient technique and, indirectly, with Planck 353 GHz polarization, and the paper claims the first 3D magnetic field characterization of diffuse H I clouds and resolved LOS variations between the LVC and IVC.

Significance. If the transfer from simulation-trained predictions to real H I clouds were established, the method would offer a genuinely new route to 3D magnetic field and turbulence diagnostics from spectroscopic 21-cm data, with potential applications to foreground subtraction and cosmic-ray studies. The paper has clear strengths: it uses an explicit conditional architecture with FiLM modulation, reports uncertainty histograms on seen and unseen simulations, includes robustness tests to noise and beam smoothing, and grounds the validation in a comparison with Planck data. However, the observational validation covers only the POS angle, and the agreement with VGT is partly internal to the same anisotropy assumption; the inferred strength, inclination, and Mach numbers for real clouds are not independently tested. The central claim therefore currently outruns the evidence.

major comments (4)
  1. [§4.2.2 and Fig. 7] The unseen-data validation rests on a single simulation with B ≈ 4 µG, σv ≈ 2.5 km/s, and γ ≈ 60°, and reports σγ dispersions near 3° with maxima near 15°, σB dispersions near 0.18 µG with maxima near 1 µG, and σMs dispersions near 0.16 with maxima near 1.0. The real-cloud conditions in Fig. 7 sit near or beyond the training grid (γ ≈ 70°, B ≈ 4.8–5.4 µG, and Ms, MA up to about 2), and no uncertainty maps are attached to Fig. 7. The claimed values of B, Ms, and MA for the LVC and IVC are therefore not demonstrated to be significant, and the single unseen simulation does not establish that the network transfers to the observed regime.
  2. [§4.3.1–4.3.2] The POS-angle validation is partly circular: both VGT and the ResNet predictions exploit the same anisotropic-MHD-turbulence assumption and both are computed from the same H I data cube. The external Planck comparison in Fig. 5 is performed on VGT integrated over the full [-40, 100] km/s range, not on the ResNet per-cloud predictions. Consequently, the AM agreement in Fig. 6 validates only the POS orientation of the network, and it does not independently test the predicted inclination, magnetic field strength, or Mach numbers.
  3. [§4.3.3] The claimed difference in magnetic field strength between the LVC (B ≈ 4.8 µG) and the IVC (B ≈ 5.4 µG) is 0.6 µG, which is only about 3σ against the quoted σB ≈ 0.18 µG dispersion and is smaller than the maximum unseen-data error of ≈ 1 µG reported in §4.2.2. No uncertainty estimate or significance test is provided for this difference, so the conclusion that the two clouds have distinct magnetic field strengths is not supported by the presented statistics.
  4. [§5.2 and §6] The independent Gaia-based stellar-anisotropy test is proposed only as future work, not performed. As submitted, the central claim in §6 that the network can reconstruct the 3D magnetic fields and turbulence parameters of diffuse H I clouds rests on the unvalidated assumption that the 27 simulation parameter sets (54 snapshots) span and encode the physical conditions of the observed LVC and IVC. This extrapolation is the weakest link in the paper and needs either an independent observational check or a substantially weakened claim.
minor comments (5)
  1. [§3.1] The phrase "thin velocity channelp" appears to contain a typographical artifact and should read "thin velocity channel maps."
  2. [§4.2.1] The word "smaler" should be "smaller," and the sentence should specify whether the quoted values are dispersions or maxima for the seen-data histograms.
  3. [§2.1] The sentence "This anisotropy is imprinted in spectroscopic observations, proving an independent way to study the magnetic fields" should read "providing an independent way," since the text is describing an opportunity rather than a proof.
  4. [Fig. 6] The AM histograms are described only qualitatively as "sharply peaked" or "concentrate near 1"; reporting the median AM and its scatter for each comparison would make the agreement quantitative and easier to compare across VGT-Planck and VGT-ResNet.
  5. [Appendix/References] The acknowledgments contain "Y .H." with an extra space, and the reference list has a formatting inconsistency in "V Y ."; these should be corrected before publication.

Circularity Check

0 steps flagged · score 0.0 of 10

No formal circularity: the network is supervised on MHD simulation ground truth, the POS-angle benchmark is anchored to Planck, and the real-cloud B/Ms/MA values are extrapolations rather than fitted renamings.

full rationale

The paper's derivation chain is not circular in the formal sense. The training labels for psi, gamma, B, M_A, and M_s are ground-truth quantities taken from the 3D MHD simulations themselves, not computed from the input channel maps by any formula that already contains the target; the network learns a mapping from synthetic spectroscopic images to simulation truth, and the fitted parameters are network weights, fitted to simulation labels, not to any real-cloud target. The observational validation in Sec. 4.3 compares the ResNet POS angle with VGT, whose integrated result is checked against Planck 353 GHz polarization, providing an external anchor for the integrated POS orientation. The real-cloud B, gamma, M_s, and M_A maps in Fig. 7 are therefore an extrapolation of a simulation-trained model rather than a claim derived from the observed data by construction. The main concerns are validation and completeness, not circularity: the per-cloud VGT comparison shares the anisotropic-MHD-turbulence premise with the training data and is not an independent measurement of B, gamma, M_s, or M_A; Sec. 5.2 explicitly defers an independent Gaia-based test to future work; and simulation and label-generation details are delegated to companion papers. These are legitimate transferability and reproducibility weaknesses, but no equation in the paper reduces a predicted quantity to an input by definition.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The paper relies on established MHD turbulence theory and the realism of multiphase simulations. No new physical entities are introduced. The main free parameters are the hand-picked simulation grid and the neural network hyperparameters, which shape the model's domain of applicability.

free parameters (2)
  • Simulation parameter grid (B, sigma_v, gamma) = B=1,3,5 uG; sigma_v=1.25,2.5,5 km/s; gamma=30,60,90 deg
    The training set is generated from 27 discrete combinations. The real clouds' physical conditions are assumed to lie within this grid. The choice is ad hoc (not derived from a unique theory) and determines the model's applicability domain.
  • Neural network hyperparameters = learning rate 2e-4, betas (0.5, 0.999), 2000 epochs (with early stopping)
    The trained model's performance depends on these choices; the paper states early stopping was monitored, so the effective number of epochs is unclear and the results may depend on the stopping criterion.
assumptions (5)
  • domain assumption MHD turbulence is anisotropic with eddies elongated along the local magnetic field (critical balance).
    The entire feature basis relies on this theory (Goldreich & Sridhar 1995; Lazarian & Vishniac 1999), used in Section 2 to link channel maps to B.
  • domain assumption The AthenaK multiphase ISM simulations reproduce realistic H I conditions.
    The DL training uses synthetic channel maps from these simulations (Section 3.1); if the simulations miss key physics, the learned mapping may not generalize to real observations.
  • domain assumption The network trained on synthetic observations generalizes to real FAST CRAFTS observations.
    The paper applies the model to real data in Section 4.3 without recalibration; the only validation is the POS angle against VGT and Planck.
  • domain assumption The velocity gradient technique (VGT) is a reliable tracer of the POS magnetic field orientation.
    Used as a benchmark in Section 4.3.1. VGT itself is based on the same anisotropy theory, though it is cross-checked with Planck polarization.
  • standard math The magnetic field strength B can be expressed as B = c_s sqrt(4*pi*rho) * M_s * M_A^{-1} (Eq. 1).
    Standard ideal MHD relation (Lazarian et al. 2022). The network is trained to predict Ms and MA, but also directly outputs B, so the relation is used for interpretation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Characterizing 3D Magnetic Fields and Turbulence in H I Clouds." pith.science (2026). https://pith.science/paper/GRYSOQTQ

@misc{pith2026250507422,
  author       = {Pith},
  title        = {Pith review of: Characterizing 3D Magnetic Fields and Turbulence in H I Clouds},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GRYSOQTQ}},
  note         = {Machine review of arXiv:2505.07422}
}
abstract

3D Galactic magnetic fields are critical for understanding the interstellar medium, Galactic foreground polarization, and the propagation of ultra-high-energy cosmic rays. Leveraging recent theoretical insights into anisotropic magnetohydrodynamic (MHD) turbulence, we introduce a deep learning framework to predict the full 3D magnetic field structure-including the plane-of-sky (POS) position angle, line-of-sight (LOS) inclination, magnetic field strength, sonic Mach number ($M_s$), and Alfv\'en Mach number ($M_A$)-from spectroscopic H~I observations. The deep learning model is trained on synthetic H~I emission data generated from multiphase 3D MHD simulations. We then apply the trained model to observational data from the Commensal Radio Astronomy FAST Survey, presenting maps of 3D magnetic field orientation, magnetic field strength, $M_s$, and $M_A$ for two H~I clouds, a low-velocity cloud (LVC) and an intermediate-velocity cloud (IVC), which overlap in the POS yet reside at different LOS distances. The deep-learning-predicted POS magnetic field position angles align closely with those determined using the velocity gradient technique, whose integrated results are consistent with independent measurements from Planck 353~GHz polarization data. This study demonstrates the potential of deep learning approaches as powerful tools for modeling the 3D distributions of 3D Galactic magnetic fields and turbulence properties throughout the Galaxy.

Figures

Figures reproduced from arXiv: 2505.07422 by the authors.

Figure 1
Figure 1. Definition of the 3D magnetic field B. B⊥ is the mag￾netic field projected on the POS and B∥ is the LOS component. ψ is B⊥’s position angle relative on the POS. γ is B’s inclination angle with respect to the LOS. ticularly, Hu et al. (2021b, 2024a) explained the correspond￾ing physics and introduced a convolutional neural network framework trained on synthetic molecular spectral line obser￾vations, reconstructing th… view at source ↗
Figure 2
Figure 2. Phase diagrams of gas number density and pressure. The black dashed line represents the thermal equilibrium obtained from Γ = Λ, where Γ and Λ are the heating and cooling functions, re￾spectively. The simulation with B ≈ 3 µG and σv ≈ 5.00 km s−1 is used. the y-axis. The simulation cubes were subsequently rotated to align the mean magnetic field inclination with respect to the LOS, or the z-axis, at angles of 90◦ , … view at source ↗
Figure 3
Figure 3. A comparison of the predicted physical quantities and the ground truth. The simulation with mean B ≈ 1 µG, σv ≈ 5.0 km s−1 , and γ ≈ 60◦ is used as an example. Panel (a): the input velocity channel map p and the corresponding spectrum. The blue shadow area indicates the velocity range used for the channel map integration. The width is 2 km s−1 . Panel (b): the predicted POS magnetic field orientation ψ, the magnetic… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Histograms of the uncertainty σ in prediction. σ is defined as the absolute difference between the predicted value and the ground truth. Panel (a): σ distributions for three simulations that are already used in the training. The three simulations all have a mean γ ≈ 60…
Figure 5
Figure 5. Figure 5: Top: A comparison between the POS magnetic field orientations derived from VGT and those obtained from Planck polarization at 353 GHz. The VGT integrates the contributions from all channels covering the velocity range of [ -40 km s−1 , 100 km s−1 ], as indicated in the…
Figure 6
Figure 6. Figure 6: Distribution functions of the AM for VGT vs. Planck (top) and VGT vs. ResNet. A positive AM corresponds to parallel alignment between the two vectors, while a negative AM represents a perpendicular alignment (see Eq. 2). to spectroscopic channels corresponding to IVC a…
Figure 7
Figure 7. Figure 7: Maps of the predicted γ, B, MA, and Ms for the LVC and IVC (see [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. 3D B-fieLds in the InterStellar medium and Star-forming regions (3D-BLISS): I. Using Starlight Polarization in the Massive IRDC Filament G11.11-0.12

    astro-ph.GA 2025-10 conditional novelty 5.0 of 10

    Starlight polarimetry toward the massive filament G11.11-0.12 yields magnetic-field inclination angles of 44-50 degrees and suggests an arc-shaped 3D field, with 3D field strengths ~80-150 microgauss.

Reference graph

Works this paper leans on

58 extracted references · 8 canonical work pages · cited by 1 Pith paper

  1. [1]

    Understanding How Encoder-Decoder Architectures Attend

    Aitken, K., Ramasesh, V . V ., Cao, Y ., & Maheswaranathan, N. 2021, arXiv e-prints, arXiv:2110.15253, doi: 10.48550/arXiv.2110.15253

    work page Pith review arXiv doi:10.48550/arxiv.2110.15253 2021

  2. [2]

    G., Lazarian, A., & Vaillancourt, J

    Andersson, B. G., Lazarian, A., & Vaillancourt, J. E. 2015, ARA&A, 53, 501, doi: 10.1146/annurev-astro-082214-122414 BICEP2 Collaboration, Ade, P. A. R., Aikin, R. W., et al. 2014, PhRvL, 112, 241101, doi: 10.1103/PhysRevLett.112.241101

    doi:10.1146/annurev-astro-082214-122414 2015

  3. [3]

    2003, MNRAS, 345, 325, doi: 10.1046/j.1365-8711.2003.06941.x

    Cho, J., & Lazarian, A. 2003, MNRAS, 345, 325, doi: 10.1046/j.1365-8711.2003.06941.x

    arXiv 2003

  4. [4]

    Cho, J., & Vishniac, E. T. 2000, ApJ, 539, 273, doi: 10.1086/309213

    doi:10.1086/309213 2000

  5. [5]

    Crutcher, R. M. 2012, ARA&A, 50, 29, doi: 10.1146/annurev-astro-081811-125514

    doi:10.1146/annurev-astro-081811-125514 2012

  6. [6]

    A., et al

    Duan, D., He, J., Bowen, T. A., et al. 2021, ApJL, 915, L8

    work page 2021

  7. [7]

    R., & Sutherland, M

    Farrar, G. R., & Sutherland, M. S. 2019, JCAP, 2019, 004, doi: 10.1088/1475-7516/2019/05/004

    doi:10.1088/1475-7516/2019/05/004 2019

  8. [8]

    Federrath, C., & Klessen, R. S. 2012, ApJ, 761, 156, doi: 10.1088/0004-637X/761/2/156 Gaia Collaboration, Vallenari, A., Brown, A. G. A., et al. 2023, A&A, 674, A1, doi: 10.1051/0004-6361/202243940

    doi:10.1088/0004-637x/761/2/156 2012

Show all 58 references
  1. [9]

    1995, ApJ, 438, 763, doi: 10.1086/175121 Gonz´alez-Casanova, D

    Goldreich, P., & Sridhar, S. 1995, ApJ, 438, 763, doi: 10.1086/175121 Gonz´alez-Casanova, D. F., & Lazarian, A. 2017, ApJ, 835, 41

    1995 doi

  2. [10]

    2022, ApJ, 934, 7, doi: 10.3847/1538-4357/ac76bf

    Ha, T., Li, Y ., Kounkel, M., et al. 2022, ApJ, 934, 7, doi: 10.3847/1538-4357/ac76bf

    2022 doi

  3. [11]

    2023, in American Astronomical Society Meeting Abstracts, V ol

    Ha, T., Li, Y ., Kounkel, M., et al. 2023, in American Astronomical Society Meeting Abstracts, V ol. 241, American Astronomical Society Meeting Abstracts, 228.01

    2023

  4. [12]

    2021, ApJL, 907, L40, doi: 10.3847/2041-8213/abd8c9

    Ha, T., Li, Y ., Xu, S., Kounkel, M., & Li, H. 2021, ApJL, 907, L40, doi: 10.3847/2041-8213/abd8c9

    2021 doi

  5. [13]

    2016, in Proceedings of the IEEE conference on computer vision and pattern recognition, 770–778

    He, K., Zhang, X., Ren, S., & Sun, J. 2016, in Proceedings of the IEEE conference on computer vision and pattern recognition, 770–778

    2016

  6. [14]

    2024, ApJ, 965, 183, doi: 10.3847/1538-4357/ad2a56

    Hoang, T., & Truong, B. 2024, ApJ, 965, 183, doi: 10.3847/1538-4357/ad2a56

    2024 doi

  7. [15]

    2025, arXiv e-prints, arXiv:2505.07423, doi: 10.48550/arXiv.2505.07423

    Hu, Y . 2025, arXiv e-prints, arXiv:2505.07423, doi: 10.48550/arXiv.2505.07423

    2025 doi arXiv

  8. [16]

    2020, Research Notes of the American Astronomical Society, 4, 105, doi: 10.3847/2515-5172/aba359 —

    Hu, Y ., & Lazarian, A. 2020, Research Notes of the American Astronomical Society, 4, 105, doi: 10.3847/2515-5172/aba359 —. 2023a, MNRAS, 519, 3736, doi: 10.1093/mnras/stac3744 —. 2023b, MNRAS, 524, 2379, doi: 10.1093/mnras/stad1996 —. 2025, ApJ, 981, 58, doi: 10.3847/1538-4357/adaf97

    2020 doi

  9. [17]

    Hu, Y ., Lazarian, A., Alina, D., Pogosyan, D., & Ho, K. W. 2023, MNRAS, 524, 2994, doi: 10.1093/mnras/stad1924

    2023 doi

  10. [18]

    2022, ApJ, 941, 92, doi: 10.3847/1538-4357/ac9df0

    Hu, Y ., Lazarian, A., Beck, R., & Xu, S. 2022, ApJ, 941, 92, doi: 10.3847/1538-4357/ac9df0

    2022 doi

  11. [19]

    2020, ApJ, 901, 162, doi: 10.3847/1538-4357/abb1c3

    Hu, Y ., Lazarian, A., Li, Y ., Zhuravleva, I., & Gendron-Marsolais, M.-L. 2020, ApJ, 901, 162, doi: 10.3847/1538-4357/abb1c3

    2020 doi

  12. [20]

    2021a, ApJ, 912, 2, doi: 10.3847/1538-4357/abedb7

    Hu, Y ., Lazarian, A., & Stanimirovi´c, S. 2021a, ApJ, 912, 2, doi: 10.3847/1538-4357/abedb7

    doi

  13. [21]

    2024a, MNRAS, 527, 11240, doi: 10.1093/mnras/stad3766

    Hu, Y ., Lazarian, A., Wu, Y ., & Fu, C. 2024a, MNRAS, 527, 11240, doi: 10.1093/mnras/stad3766

    doi

  14. [22]

    2021b, ApJ, 915, 67, doi: 10.3847/1538-4357/ac00ab

    Hu, Y ., Lazarian, A., & Xu, S. 2021b, ApJ, 915, 67, doi: 10.3847/1538-4357/ac00ab

    doi

  15. [23]

    M., & Lazarian, A

    Hu, Y ., Xu, S., Arzamasskiy, L., Stone, J. M., & Lazarian, A. 2024b, MNRAS, 527, 3945, doi: 10.1093/mnras/stad3493

    doi

  16. [24]

    2021c, ApJ, 911, 37, doi: 10.3847/1538-4357/abea18 —

    Hu, Y ., Xu, S., & Lazarian, A. 2021c, ApJ, 911, 37, doi: 10.3847/1538-4357/abea18 —. 2021d, ApJ, 911, 37, doi: 10.3847/1538-4357/abea18

    doi

  17. [25]

    H., & Lazarian, A

    Hu, Y ., Yuen, K. H., & Lazarian, A. 2018, MNRAS, 480, 1333, doi: 10.1093/mnras/sty1807

    2018 doi

  18. [26]

    H., Lazarian, V ., et al

    Hu, Y ., Yuen, K. H., Lazarian, V ., et al. 2019, Nature Astronomy, 3, 776, doi: 10.1038/s41550-019-0769-0

    2019 doi

  19. [27]

    2016, MNRAS, 461, 1227, doi: 10.1093/mnras/stw1296

    Kandel, D., Lazarian, A., & Pogosyan, D. 2016, MNRAS, 461, 1227, doi: 10.1093/mnras/stw1296

    2016 doi

  20. [28]

    2018, arXiv e-prints, arXiv:1805.07544, doi: 10.48550/arXiv.1805.07544

    Kang, B., Lijffijt, J., & De Bie, T. 2018, arXiv e-prints, arXiv:1805.07544, doi: 10.48550/arXiv.1805.07544

    2018 doi arXiv

  21. [29]

    2010, ApJ, 720, 742, doi: 10.1088/0004-637X/720/1/742 10

    Kowal, G., & Lazarian, A. 2010, ApJ, 720, 742, doi: 10.1088/0004-637X/720/1/742 10

    2010 doi

  22. [30]

    2002, ApJL, 564, L97, doi: 10.1086/338978

    Koyama, H., & Inutsuka, S.-i. 2002, ApJL, 564, L97, doi: 10.1086/338978

    2002 doi

  23. [31]

    Larson, R. B. 1981, MNRAS, 194, 809, doi: 10.1093/mnras/194.4.809

    1981 doi

  24. [32]

    2007, JQSRT, 106, 225, doi: 10.1016/j.jqsrt.2007.01.038

    Lazarian, A. 2007, JQSRT, 106, 225, doi: 10.1016/j.jqsrt.2007.01.038

    2007 doi

  25. [33]

    2000, ApJ, 537, 720, doi: 10.1086/309040

    Lazarian, A., & Pogosyan, D. 2000, ApJ, 537, 720, doi: 10.1086/309040

    2000 doi

  26. [34]

    Lazarian, A., & Vishniac, E. T. 1999, ApJ, 517, 700, doi: 10.1086/307233

    1999 doi

  27. [35]

    Lazarian, A., & Yuen, K. H. 2018, ApJ, 853, 96, doi: 10.3847/1538-4357/aaa241

    2018 doi

  28. [36]

    H., & Pogosyan, D

    Lazarian, A., Yuen, K. H., & Pogosyan, D. 2022, ApJ, 935, 77, doi: 10.3847/1538-4357/ac6877 —. 2024, ApJ, 974, 237, doi: 10.3847/1538-4357/ad6d62

    2022 doi

  29. [37]

    2018, IEEE Microwave Magazine, 19, 112, doi: 10.1109/MMM.2018.2802178

    Li, D., Wang, P., Qian, L., et al. 2018, IEEE Microwave Magazine, 19, 112, doi: 10.1109/MMM.2018.2802178

    2018

  30. [38]

    2022, MNRAS, 510, 4952, doi: 10.1093/mnras/stab3783 Mac Low, M.-M., & Klessen, R

    Liu, M., Hu, Y ., & Lazarian, A. 2022, MNRAS, 510, 4952, doi: 10.1093/mnras/stab3783 Mac Low, M.-M., & Klessen, R. S. 2004, Reviews of Modern Physics, 76, 125, doi: 10.1103/RevModPhys.76.125

    2022 doi

  31. [39]

    2001, ApJ, 554, 1175, doi: 10.1086/321413

    Maron, J., & Goldreich, P. 2001, ApJ, 554, 1175, doi: 10.1086/321413

    2001 doi

  32. [40]

    2020, Frontiers in Astronomy and Space Sciences, 7, 83

    Matteini, L., Franci, L., Alexandrova, O., et al. 2020, Frontiers in Astronomy and Space Sciences, 7, 83

    2020

  33. [41]

    F., & Ostriker, E

    McKee, C. F., & Ostriker, E. C. 2007, ARA&A, 45, 565, doi: 10.1146/annurev.astro.45.051806.110602

    2007 arXiv

  34. [42]

    F., & Ostriker, J

    McKee, C. F., & Ostriker, J. P. 1977, ApJ, 218, 148, doi: 10.1086/155667

    1977 doi

  35. [43]

    1965, QJRAS, 6, 265

    Mestel, L. 1965, QJRAS, 6, 265

    1965

  36. [44]

    2012, A&A, 542, A93, doi: 10.1051/0004-6361/201118526

    Oppermann, N., Junklewitz, H., Robbers, G., et al. 2012, A&A, 542, A93, doi: 10.1051/0004-6361/201118526

    2012 doi

  37. [45]

    V ., Markopoulioti, L., Bouzelou, F., et al

    Panopoulou, G. V ., Markopoulioti, L., Bouzelou, F., et al. 2025, ApJS, 276, 15, doi: 10.3847/1538-4365/ad8b21

    2025 doi

  38. [46]

    2017, arXiv e-prints, arXiv:1709.07871, doi: 10.48550/arXiv.1709.07871 Planck Collaboration, Adam, R., Ade, P

    Perez, E., Strub, F., de Vries, H., Dumoulin, V ., & Courville, A. 2017, arXiv e-prints, arXiv:1709.07871, doi: 10.48550/arXiv.1709.07871 Planck Collaboration, Adam, R., Ade, P. A. R., et al. 2016, A&A, 586, A133, doi: 10.1051/0004-6361/201425034 Planck Collaboration, Akrami, ...

    2017 doi arXiv

  39. [47]

    2024, arXiv e-prints, arXiv:2411.11157, doi: 10.48550/arXiv.2411.11157

    Schmaltz, T., Hu, Y ., & Lazarian, A. 2024, arXiv e-prints, arXiv:2411.11157, doi: 10.48550/arXiv.2411.11157

    2024 doi arXiv

  40. [48]

    M., Mullen, P

    Stone, J. M., Mullen, P. D., Fielding, D., et al. 2024, arXiv e-prints, arXiv:2409.16053, doi: 10.48550/arXiv.2409.16053

    2024 doi arXiv

  41. [49]

    C., Soler, J

    Tahani, M., Plume, R., Brown, J. C., Soler, J. D., & Kainulainen, J. 2019, A&A, 632, A68, doi: 10.1051/0004-6361/201936280 Telescope Array Collaboration, Abbasi, R. U., Allen, M. G., et al. 2023, Science, 382, 903, doi: 10.1126/science.abo5095

    2019 doi

  42. [50]

    Unger, M., & Farrar, G. R. 2025, arXiv e-prints, arXiv:2502.15876, doi: 10.48550/arXiv.2502.15876 Van Rossum, G., & Drake, F. L. 2009, Python 3 Reference Manual (Scotts Valley, CA: CreateSpace) V´azquez-Semadeni, E., Gazol, A., & Scalo, J. 2000, ApJ, 540, 271, doi: 10.1086/309318

    2025 doi arXiv

  43. [51]

    2016, ApJ, 816, 15, doi: 10.3847/0004-637X/816/1/15

    Wang, X., Tu, C., Marsch, E., He, J., & Wang, L. 2016, ApJ, 816, 15, doi: 10.3847/0004-637X/816/1/15

    2016 doi

  44. [52]

    G., Hollenbach, D., McKee, C

    Wolfire, M. G., Hollenbach, D., McKee, C. F., Tielens, A. G. G. M., & Bakes, E. L. O. 1995, ApJ, 443, 152, doi: 10.1086/175510

    1995 doi

  45. [53]

    G., McKee, C

    Wolfire, M. G., McKee, C. F., Hollenbach, D., & Tielens, A. G. G. M. 2003, ApJ, 587, 278, doi: 10.1086/368016

    2003 doi

  46. [54]

    2025, ApJ, 980, 52, doi: 10.3847/1538-4357/ada8a0

    Xu, D., Karcheski, J., Law, C.-Y ., et al. 2025, ApJ, 980, 52, doi: 10.3847/1538-4357/ada8a0

    2025 doi

  47. [55]

    2020, ApJ, 894, 63, doi: 10.3847/1538-4357/ab8465

    Xu, S., & Lazarian, A. 2020, ApJ, 894, 63, doi: 10.3847/1538-4357/ab8465

    2020 doi

  48. [56]

    H., Ho, K

    Yuen, K. H., Ho, K. W., & Lazarian, A. 2021, ApJ, 910, 161, doi: 10.3847/1538-4357/abe4d4

    2021 doi

  49. [57]

    2024, arXiv e-prints, arXiv:2411.07080, doi: 10.48550/arXiv.2411.07080

    Zhang, J., Hu, Y ., & Lazarian, A. 2024, arXiv e-prints, arXiv:2411.07080, doi: 10.48550/arXiv.2411.07080

    2024 doi arXiv

  50. [58]

    Z., Yuen, K

    Zhao, S., Yan, H., Liu, T. Z., Yuen, K. H., & Wang, H. 2024, Nature Astronomy, 8, 725, doi: 10.1038/s41550-024-02249-0

    2024 doi

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.