Pith. sign in

REVIEW 5 major objections 4 minor 2 cited by

A fast and robust recipe for modeling non-ideal MHD effects in star-formation simulations

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

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

Pith's one-line read Non-ideal MHD resistivities in star-formation simulations can be computed from an evolving power-law lookup table, eliminating the chemical network.

desk verdict A useful surrogate for non-ideal MHD resistivities, with honest blind tests and public code, but the single-coordinate interpolation may not survive contact with turbulent multi-core clouds. read the letter →

arxiv 2502.06933 v1 pith:6BKEKSIE submitted 2025-02-10 astro-ph.GA

E. Agianoglou , A. Tritsis , K. Tassis This is my paper · ORCID
classification astro-ph.GA
keywords non-idealMHDambipolardiffusionstarformationsimulationsionizationfractionresistivitieschemicalnetworkinterpolationmolecularclouds
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

The paper tries to establish that ambipolar diffusion and other non-ideal MHD effects, which control whether magnetic fields slow or stop star-forming collapse, can be modeled without solving a chemical network on every grid cell. The authors reduce the resistivities to two ingredients: a total ion number density and a nearly constant cloud-chemistry factor. They approximate the ion abundance with a power law in density that evolves as the cloud collapses, with the power-law coefficients interpolated from just eight chemodynamical simulations. When benchmarked against full non-ideal MHD runs with non-equilibrium chemistry, including blind 2D tests and a 3D turbulent run, the approximation reproduces collapse timescales, morphology, and magnetic-field evolution while speeding up the calculation by a factor of 100 to 1000. The method targets densities up to about $10^{6}$ $cm^{-3}$ and is aimed at pc-scale molecular cloud simulations.

What carries the argument

The central object is the evolving power-law fit to the total ion abundance, chi_i,tot = A (rho/rho_int)^B, with rho_int the geometric mean of the maximum and log-mean H2 density. Its coefficient A and exponent B are interpolated in each grid cell from tabulated values generated by eight chemodynamical models, using the ad hoc functional forms given in Appendix B, and the cloud-chemistry constant C_perp is multi-linearly interpolated from the same models. These two interpolated quantities feed the simplified resistivity formulas, replacing the chemical network entirely.

What would settle it

Run a non-ideal MHD simulation with the public tables in a regime where the power law is known to struggle: early collapse at central densities below $10^{4}$ $cm^{-3}$, where the paper's Appendix A says a single power law underpredicts the ion abundance by up to two orders of magnitude. If the interpolated chi_i,tot differs from a full chemical network by the same factor at those early times, the method's collapse-time accuracy at later epochs would need to be re-examined.

Watch

Extended reading notes

Core claim

The core claim is that for molecular cloud conditions below about $10^{6}$ $cm^{-3}$, the resistivities entering the generalized Ohm law can be written as eta_parallel approximately C_parallel rho_H2 / n_i,tot and eta_perp approximately C_perp $v_A^{2}$ / n_i,tot, with the Hall resistivity negligible, so that knowing the total ion density n_i,tot suffices. The paper further claims that n_i,tot is accurately described by the evolving power law chi_i,tot = A (rho/rho_int)^B, where A and B are interpolated from eight fiducial chemodynamical models as functions of local physical conditions, and that the constant C_perp is nearly independent of the cloud's evolution. Using this approximation in 2D and 3D non-ideal MHD simulations, the authors report excellent agreement with full non-equilibrium chemistry runs for the time to reach a central density of $10^{6}$ $cm^{-3}$, the spatial structure and fragmentation of the cloud, and the B-rho relation, at a computational speedup of two to four orders of magnitude.

Load-bearing premise

The load-bearing premise is that the total ion abundance in each cell of a simulated cloud is always well described by the evolving power law chi_i,tot = A (rho/rho_int)^B, with A and B interpolated from just eight chemodynamical models through the ad hoc functional forms in Appendix B; if the abundance ever departs from this power-law form, the resistivities and the resulting collapse are wrong.

Editorial extensions

If this is right

  • Cloud-collapse runs that would otherwise require a chemical network on every grid cell can instead use the public table and Fortran interpolation, cutting the resistivity cost by a factor of roughly 100 in 2D and up to roughly 10,000 in 3D.
  • The method reproduces ambipolar-diffusion timescales, core morphology, and the magnetic field-versus-density relation in 2D and in a 3D turbulent run with random initial conditions, so pc-scale simulations can now explore a large parameter space.
  • The approach is valid up to number densities of about 10^6 cm^-3, covering the fragmentation and core-formation phase before sink particles form at about 10^7 cm^-3.
  • Because the ion-abundance fit evolves with the cloud, it avoids the single-power-law bias that the paper documents in Appendix A for early-time abundances.
  • Other approximate schemes in the literature (Shu 1992; Tassis et al. 2012b; Tielens 2005; Tsukamoto et al. 2022) either delay collapse by several Myr or fail to collapse at all in the benchmark shown in Fig. 6.

Reading between the lines

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

  • Extension left implicit: the same tabulated-power-law strategy could be carried to higher densities by adding grain charge and grain-size distribution as extra interpolation axes, since the paper's 10^6 cm^-3 ceiling comes specifically from ignoring grains.
  • A cheap stress test the authors did not run: take one of the public tables and evaluate chi_i,tot against the chemical network inside a sheared or shocked region where the density distribution may not be lognormal; if the power-law fit fails there, the method's usefulness in turbulent boxes depends on the specific density estimator used.
  • Related consequence: because C_perp is assumed constant per cloud, the method implicitly assumes the dominant ion mass does not change along a collapse; probing chemistries where the dominant ion switches early (for example from C+ to HCO+ or H3+) would tell whether a per-species table is needed.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 4 minor

Summary. The paper presents an empirical approximation for computing non-ideal MHD resistivities in star-formation simulations without evolving a chemical network. Ion abundances are represented by an evolving power law chi_i,tot = A (rho_H2 / rho_int)^B (Eq. A.1), with A, B, and the perpendicular-conductivity constant C_perp interpolated from eight 2D axisymmetric chemodynamical collapse models as functions of cosmic-ray ionization rate, visual extinction, temperature, and initial density, using the ad hoc forms in Eqs. B.5-B.8. The resistivities follow from simplified physical expressions (Eq. 11), with Hall resistivity set to zero. The method is benchmarked against the training simulations, against additional blind 2D models with random initial conditions, against a 3D turbulent simulation, and against earlier approximate methods. The authors report a speed-up factor of 100-1000 and provide tabulated data and a Fortran implementation.

Significance. If the method holds up, it would be practically valuable: it offers a cheap route to including ambipolar diffusion in pc-scale 3D simulations without a chemical network, potentially enabling wider parameter-space exploration in star-formation studies. The paper's strengths include the physically motivated derivation of the simplified resistivity expressions, the explicit blind tests in Fig. 2, the inclusion of a 3D test with turbulence, the comparison to existing approximations in Fig. 6, and the public release of the interpolation code and data. The main weaknesses are that the quantitative accuracy claims rest on visual comparisons, the one-scalar-coordinate (rho_int) description of chemical evolution is validated only indirectly, and the interpolation formulas are not cross-validated. These are addressable but need to be fixed before the claimed accuracy range and generalizability are established.

major comments (5)
  1. [§4.2, Figs. 2 and 5] The central claim of "excellent quantitative and qualitative agreement" is not backed by any numerical error statistic. Fig. 2 and the 3D test in Fig. 5 are described visually, but the manuscript reports no quantitative deviation in collapse time, central-density trajectory, density-field difference, or fragmentation morphology. Please report concrete metrics, for example relative errors in the time to reach a given central density, maximum density differences at matched times, and a cell-by-cell or core-by-core comparison of the 3D run. This is needed to separate a real predictive success from agreement that is merely plausible by eye.
  2. [Appendix A, Eq. A.1; §4.5] The method assumes that the total ion abundance throughout a cloud is always a single power law in local density, with coefficients A and B set by the single global scalar rho_int. In a turbulent, multi-core 3D cloud, cells with the same n_H2, T, Av, and zeta but different chemical histories are assigned identical chi_i, even though non-equilibrium chemistry can yield different values. The paper itself notes (Appendix A, Fig. A.1) that a non-evolving power law underpredicts the early-time abundance by up to two orders of magnitude, which shows how sensitive chi_i is to the evolutionary state. The only 3D evidence is one central-slice image at one time (Fig. 5). Please add a quantitative assessment of this single-coordinate assumption, for example by comparing predicted versus chemical-network ion abundances cell-by-cell in the 3D run, and by including a test with multiple cores at different collapse stages.
  3. [Appendix B, Eqs. B.5-B.8] No leave-one-out cross-validation of the interpolation function is reported. Equations B.5-B.8 are ad hoc functional forms with several structural parameters, and Fig. 1 is a training-set check, not an independent benchmark. The blind models in Fig. 2 provide some independent evidence, but they cover only a handful of points in a five-dimensional parameter space. Please perform a leave-one-out test over the eight interpolation models and quote the resulting interpolation errors in A, B, and in the derived resistivities as a function of the physical parameters.
  4. [Abstract vs. §6] The abstract states a factor of 100-1000 increase in computational speed, while Section 6 reports a factor of about 10^2 in 2D and 10^4 in 3D. This is a central quantitative claim of the paper, so the discrepancy should be resolved with a precise statement of how the timings were measured and what exactly is being compared (e.g., wall-clock time per simulation, total cost including interpolation setup, or cost per hydrodynamical step).
  5. [§2.3 and §5.2] The validity range up to n_H2 ~ 10^6 cm^-3 is asserted despite two known sources of error at the upper end: the neglect of the Hall term, which the authors estimate can introduce 25% error in the last half dex, and the neglect of grains, which some earlier work finds can dominate the perpendicular resistivity above ~10^4 cm^-3 depending on the grain-size distribution. The 3D test is reassuring, but it is compared only visually and includes grains in the full-chemistry simulation. Please quantify how large the resistivity errors can become within 10^5-10^6 cm^-3 under the assumptions made, and show whether the resulting dynamical differences remain within the claimed accuracy.
minor comments (4)
  1. [Appendix B, Eq. B.4] The pseudo-code in Eq. B.4 would be easier to follow if the selection logic for f_zeta, f_Av, and the corresponding model columns were written consistently for all three variables, including the case where the requested value is exactly equal to the fiducial value.
  2. [Appendix A] The text calls rho_int the geometric mean of the maximum density and the "logarithmic mean" of the H2 density, but the displayed definition rho_int = sqrt(rho_max * 10^<log10 rho>) is the geometric mean of the maximum and the geometric-mean density. Please unify the terminology.
  3. [References] The reference list includes Mouschovias (1995), but I did not find a corresponding citation in the text; please add the citation or remove the reference.
  4. [Fig. 1 and Fig. 2] For the reader's benefit, the figure captions should explicitly note that Fig. 1 uses models that entered the interpolation, while Fig. 2 uses models that did not, even though this distinction is discussed in the text.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the interpolation is explicitly calibrated on the eight labeled training models, and the paper's key claims are tested on separate blind 2D and 3D simulations.

full rationale

The derivation chain is self-contained and does not reduce to its own inputs. The resistivity formulas (Eqs. 1-11) are obtained from standard conductivity theory, with the only empirical ingredient being the explicitly stated power-law ansatz chi_i,tot = A (rho/rho_int)^B (Eq. A.1). The coefficients A and B are tabulated from the eight chemodynamical simulations listed in the first half of Table 1, and the paper is transparent about this: Fig. 1 is labeled 'models used in the interpolation', and Section 4.2 states that 'results from these simulations were already present in our interpolating function'. The independent evidence for the method comes from the 'Generalized Models' tested blindly (bottom half of Table 1, shown in Fig. 2) and from the 3D turbulent simulation with non-equilibrium chemistry (Fig. 5). These blind tests use physical conditions not used in constructing the interpolation, including a model with random initial conditions and a supercritical mass-to-flux ratio, so the central claim is not merely a refit of the training data. The self-citations to Tritsis et al. (2022, 2023) provide the reference chemical network and numerical setup, but they are not invoked as an unverified uniqueness theorem or as the source of the target result. The concern raised about whether a single global rho_int can encode the local non-equilibrium chemical state in multi-core turbulent clouds is a legitimate validation/extrapolation risk, but it is not circularity: the paper's approximation could be wrong in that regime without being equivalent to its inputs by construction. Overall, no circular step is exhibited in the paper's own equations or claims.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The method's accuracy rests on a tabulated, ad hoc interpolation of ion abundances and resistivities fitted from a small set of full-chemistry simulations. The physical resistivity formulas themselves are standard, but the interpolation surface, its functional form, and the grain-free assumption are the parts a reader must accept without independent derivation.

free parameters (3)
  • A and B coefficients of the power-law ionization fraction (Eq. A.1) = Tabulated per model as functions of rho_int; values provided in the GitHub table
    Fitted to the chemodynamical simulations in the top half of Table 1; the interpolation surface is built from these coefficients.
  • C_perp constant (Table B.1) = 7.0 to 7.8 x 10^-12 cm^-5 s^3
    Fitted per model from the chemo simulations and used in Eq. 11 for the perpendicular resistivity; it is nearly constant but still model-dependent.
  • Interpolation structural parameters (f_zeta, f_Av, f_T, alpha, beta, expAv) = For example, f_zeta = 2 or 0.5; alpha and beta derived from A ratios
    Hand-chosen functional forms in Eqs. B.5-B.8 that force the interpolation through the training models; they are not derived from chemistry.
assumptions (6)
  • standard math Standard generalized Ohm's law and conductivity definitions (Parks 1991; Kunz & Mouschovias 2009) are valid.
    Used in Section 2, Eqs. 1-4, without proof.
  • domain assumption For n_H2 up to 10^9 cm^-3, ions and electrons are strongly magnetized: omega_s tau_sn >> 1.
    Invoked in Section 2.2 to simplify the perpendicular and Hall conductivities; supported by cited earlier work.
  • domain assumption Charge neutrality n_e = n_i,tot holds everywhere in the cloud.
    Used to set the Hall conductivity to zero in Eq. 10 and to express the parallel conductivity through n_i,tot.
  • ad hoc to paper The total ion abundance follows a single evolving power law in H2 density: chi_i,tot = A (rho_H2 / rho_int)^B.
    Eq. A.1 is an empirical ansatz; the paper acknowledges it underpredicts early-time abundances by up to two orders of magnitude.
  • domain assumption The interpolation surface built from eight chemo models (Table 1 top) generalizes across the target parameter space, including mass-to-flux ratios not used as interpolation variables.
    This is central to the method; it is tested only on a handful of blind models.
  • domain assumption Grains do not affect the resistivities significantly below n_H2 ~ 10^6 cm^-3.
    Section 5.2 bases this on one grainy 2D model and cited literature; the validity is conditional on the grain size distribution.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A fast and robust recipe for modeling non-ideal MHD effects in star-formation simulations." pith.science (2026). https://pith.science/paper/6BKEKSIE

@misc{pith2026250206933,
  author       = {Pith},
  title        = {Pith review of: A fast and robust recipe for modeling non-ideal MHD effects in star-formation simulations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6BKEKSIE}},
  note         = {Machine review of arXiv:2502.06933}
}
read the original abstract

Non-ideal MHD effects are thought to be a crucial component of the star-formation process. Numerically, several complications render the study of non-ideal MHD effects in 3D simulations extremely challenging and hinder our efforts of exploring a large parameter space. We aim to overcome such challenges by proposing a novel, physically-motivated empirical approximation to model non-ideal MHD effects. We perform a number of 2D axisymmetric 3-fluid non-ideal MHD simulations of collapsing prestellar cores and clouds with non-equilibrium chemistry and leverage upon previously-published results. We utilize these simulations to develop a multivariate interpolating function to predict the ionization fraction in each region of the cloud depending on the local physical conditions. We subsequently use analytically-derived, simplified expressions to calculate the resistivities of the cloud in each grid cell. Therefore, in our new approach the resistivities are calculated without the use of a chemical network. We benchmark our method against additional 2D axisymmetric non-ideal MHD simulations with random initial conditions and a 3D non-ideal MHD simulation with non-equilibrium chemistry. We find excellent quantitative and qualitative agreement between our approach and the "full" non-ideal MHD simulations both in terms of the spatial structure of the simulated clouds and regarding their time evolution. We achieve a factor of 100-1000 increase in computational speed. Given that we ignore the contribution of grains, our approximation is valid up to number densities of 10^6 cm^(-3) and is therefore suitable for pc-scale simulations of molecular clouds. The tabulated data required for integrating our method in hydrodynamical codes, along with a fortran implementation of the interpolating function are publicly available at https://github.com/manosagian/Non-Ideal-MHD-Approximate-Code.

Figures

Figures reproduced from arXiv: 2502.06933 by the authors.

Figure 1
Figure 1. Comparisons of the time evolution of the central H2 number density for models used in the interpolation. The black lines represent simulations using a full chemical network and the red lines represent simulations using our approximate method. Different linestyles are used to denote models with different physical parameters (see legend). The green line represents a model with the same physical parameters as the Fiduc… view at source ↗
Figure 2
Figure 2. Same as [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. The magnetic field at the center of the cloud as a func￾tion of the central H2 number density for the fiducial and ran￾dom ICs models. The black and blue lines represent the fiducial and random ICs simulations respectively using a full chemical network. The red and purple lines represent the fiducial and ran￾dom ICs simulations respectively using the method described in this paper We can see that the function approa… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Comparison of the spatial distribution of the density between a simulation where the resistivities are calculated using our non-equilibrium chemical network (top row) and a simulation where the resistivities are computed using our approximation (bottom row) for our fid…
Figure 5
Figure 5. Figure 5: H2 density on the central x-y slice for a chemodynamical 3D simulation (top) and a simulation using our method (bottom) at 2.5 Myrs simulation time. The two simulations seem to have evolved very similarly, with only minor differences. The bottom simulation has collapse…
Figure 6
Figure 6. Figure 6: Central H2 number density evolution in simulations with identical initial conditions (fiducial model; see [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. Accretion across scales: streamers, surface-layer transport, and rapid replenishment in young protoplanetary discs

    astro-ph.EP 2026-08 conditional novelty 6.0 of 10

    Cloud-fed ideal-MHD zoom-in simulations of nine young stars show discs are replenished on ~10,000-year timescales via surface-layer accretion and can be truncated by massive streamers.

  2. Ambipolar diffusion and the mass-to-flux ratio in a turbulent collapsing cloud

    astro-ph.GA 2025-05 conditional novelty 6.0 of 10

    In a turbulent collapsing cloud, ambipolar drift becomes chaotic and antiphased above and below the midplane, but the region-averaged true mass-to-flux ratio still increases monotonically, while observed values do not...

Reference graph

Works this paper leans on

46 extracted references · 10 canonical work pages · cited by 2 Pith papers

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archiveprefix author booktitle chapter edition editor howpublished institution eprint journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.all := #1 ...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in " " * FUNCTION format....

  3. [3]

    ichiro .\ 2024, , 961, 100

    Abe, D., Inoue, T., & Inutsuka, S.-. ichiro .\ 2024, , 961, 100. doi:10.3847/1538-4357/ad072a

    doi:10.3847/1538-4357/ad072a 2024

  4. [4]

    M., Burkhart, B., Semenov, V

    Appel, S. M., Burkhart, B., Semenov, V. A., et al.\ 2023, , 954, 93. doi:10.3847/1538-4357/ace897

    doi:10.3847/1538-4357/ace897 2023

  5. [5]

    E., Dapp, W

    Basu, S., Ciolek, G. E., Dapp, W. B., et al.\ 2009, , 14, 483. doi:10.1016/j.newast.2009.01.004

    doi:10.1016/j.newast.2009.01.004 2009

  6. [6]

    M., Terzieva, R., et al.\ 1998, , 499, 234

    Caselli, P., Walmsley, C. M., Terzieva, R., et al.\ 1998, , 499, 234. doi:10.1086/305624

    doi:10.1086/305624 1998

  7. [7]

    Clark, P. C. & Glover, S. C. O.\ 2014, , 444, 2396. doi:10.1093/mnras/stu1589

    work page doi:10.1093/mnras/stu1589 2014

  8. [8]

    doi:10.1007/BF01448839

    Courant, R., Friedrichs, K., & Lewy, H.\ 1928, Mathematische Annalen, 100, 32. doi:10.1007/BF01448839

    doi:10.1007/bf01448839 1928

Show all 46 references
  1. [9]

    Desch, S. J. & Mouschovias, T. C.\ 2001, , 550, 314. doi:10.1086/319703

    2001 doi

  2. [10]

    G.\ 1979, , 232, 729

    Elmegreen, B. G.\ 1979, , 232, 729. doi:10.1086/157333

    1979 doi

  3. [11]

    R., Millman, K

    Harris, C. R., Millman, K. J., van der Walt, S. J. et al.\ 2020 Nature 585, 357–362. doi: 0.1038/s41586-020-2649-2

    2020

  4. [12]

    D.\ 2007, Computing in Science & Engineering, vol

    Hunter, J. D.\ 2007, Computing in Science & Engineering, vol. 9, no. 3, pp. 90-95

    2007

  5. [13]

    doi:10.1093/pasj/psx089

    Inoue, T., Hennebelle, P., Fukui, Y., et al.\ 2018, , 70, S53. doi:10.1093/pasj/psx089

    2018 doi

  6. [14]

    R., Federrath, C., et al.\ 2019, , 488, 1407

    Khullar, S., Krumholz, M. R., Federrath, C., et al.\ 2019, , 488, 1407. doi:10.1093/mnras/stz1800

    2019 doi

  7. [15]

    Kunz, M. W. & Mouschovias, T. C.\ 2009, , 693, 1895. doi:10.1088/0004-637X/693/2/1895

    2009 doi

  8. [16]

    Kunz, M. W. & Mouschovias, T. C.\ 2010, , 408, 322. doi:10.1111/j.1365-2966.2010.17110.x

    2010

  9. [17]

    K., Pitrou A., & Seibert S

    Lam S. K., Pitrou A., & Seibert S. 2015. Numba: A LLVM-based Python JIT compiler. In Proceedings of the 2nd Workshop on the LLVM Compiler Infrastructure in HPC (LLVM'15). ACM, New York, NY. doi: 10.1145/2833157.2833162

    2015

  10. [18]

    doi:10.1051/0004-6361/201424081

    Lef \`e vre, C., Pagani, L., Juvela, M., et al.\ 2014, , 572, A20. doi:10.1051/0004-6361/201424081

    2014 doi

  11. [19]

    S., McKee, C

    Li, P. S., McKee, C. F., & Klein, R. I.\ 2006, , 653, 1280. doi:10.1086/508977

    2006 doi

  12. [20]

    L., Konigl, A., et al.\ 1995, , 442, 726

    Mac Low, M.-M., Norman, M. L., Konigl, A., et al.\ 1995, , 442, 726. doi:10.1086/175477

    1995 doi

  13. [21]

    doi:10.1051/0004-6361/201526780

    Marchand, P., Masson, J., Chabrier, G., et al.\ 2016, , 592, A18. doi:10.1051/0004-6361/201526780

    2016 doi

  14. [22]

    doi:10.1051/0004-6361/202244291

    Marchand, P., Lebreuilly, U., Mac Low, M.-M., et al.\ 2023, , 670, A61. doi:10.1051/0004-6361/202244291

    2023 doi

  15. [23]

    C.\ 1995, The Physics of the Interstellar Medium and Intergalactic Medium, 80, 184

    Mouschovias, T. C.\ 1995, The Physics of the Interstellar Medium and Intergalactic Medium, 80, 184

    1995

  16. [24]

    Mouschovias, T. C. & Ciolek, G. E.\ 1999, The Origin of Stars and Planetary Systems, 540, 305

    1999

  17. [25]

    Mouschovias, T. C. & Spitzer, L.\ 1976, , 210, 326. doi:10.1086/154835

    1976 doi

  18. [26]

    doi:10.1086/340587

    Nakano, T., Nishi, R., & Umebayashi, T.\ 2002, , 573, 199. doi:10.1086/340587

    2002 doi

  19. [27]

    K.\ 1991, Redwood City, CA, Addison-Wesley Publishing Co., 1991

    Parks, G. K.\ 1991, Redwood City, CA, Addison-Wesley Publishing Co., 1991

    1991

  20. [28]

    doi:10.1051/0004-6361:20078818

    Pinto, C., Galli, D., & Bacciotti, F.\ 2008, , 484, 1. doi:10.1051/0004-6361:20078818

    2008 doi

  21. [29]

    & Galli, D.\ 2008, , 484, 17

    Pinto, C. & Galli, D.\ 2008, , 484, 17. doi:10.1051/0004-6361:20078819

    2008 doi

  22. [30]

    H.\ 1983, , 273, 202

    Shu, F. H.\ 1983, , 273, 202. doi:10.1086/161359

    1983 doi

  23. [31]

    H., Adams, F

    Shu, F. H., Adams, F. C., & Lizano, S.\ 1987, , 25, 23. doi:10.1146/annurev.aa.25.090187.000323

    1987

  24. [32]

    H.\ 1992, The physics of astrophysics

    Shu, F. H.\ 1992, The physics of astrophysics. Volume II: Gas dynamics., by Shu, F. H.. University Science Books, Mill Valley, CA (USA), 1992, 493 p., ISBN 0-935702-65-2, Price 36.95

    1992

  25. [33]

    doi:10.1051/0004-6361/201425434

    Steinacker, J., Andersen, M., Thi, W.-F., et al.\ 2015, , 582, A70. doi:10.1051/0004-6361/201425434

    2015 doi

  26. [34]

    & Mouschovias, T

    Tassis, K. & Mouschovias, T. C.\ 2007, , 660, 388. doi:10.1086/512761

    2007 doi

  27. [35]

    W., et al.\ 2012a, , 753, 29

    Tassis, K., Willacy, K., Yorke, H. W., et al.\ 2012a, , 753, 29. doi:10.1088/0004-637X/753/1/29

    doi

  28. [36]

    W., et al.\ 2012b, , 754, 6

    Tassis, K., Willacy, K., Yorke, H. W., et al.\ 2012b, , 754, 6. doi:10.1088/0004-637X/754/1/6

    doi

  29. [37]

    Tielens, A. G. G. M.\ 2005, The Physics and Chemistry of the Interstellar Medium, by A. G. G. M. Tielens, pp. . ISBN 0521826349. Cambridge, UK: Cambridge University Press, 2005

    2005

  30. [38]

    V., Mouschovias, T

    Tritsis, A., Panopoulou, G. V., Mouschovias, T. C., et al.\ 2015, , 451, 4384. doi:10.1093/mnras/stv1133

    2015 doi

  31. [39]

    doi:10.1093/mnras/stab3740

    Tritsis, A., Federrath, C., Willacy, K., et al.\ 2022, , 510, 4420. doi:10.1093/mnras/stab3740

    2022 doi

  32. [40]

    doi:10.1093/mnras/stad829

    Tritsis, A., Basu, S., & Federrath, C.\ 2023, , 521, 5087. doi:10.1093/mnras/stad829

    2023 doi

  33. [41]

    doi:10.1093/mnras/stv1290

    Tsukamoto, Y., Iwasaki, K., Okuzumi, S., et al.\ 2015, , 452, 278. doi:10.1093/mnras/stv1290

    2015 doi

  34. [42]

    doi:10.48550/arXiv.2209.13765

    Tsukamoto, Y., Maury, A., Commer c on, B., et al.\ 2022, arXiv:2209.13765. doi:10.48550/arXiv.2209.13765

    2022 doi arXiv

  35. [43]

    J., Smith, B

    Turk, M. J., Smith, B. D., Oishi, J. S., et al.\ 2011, , 192, 9. doi:10.1088/0067-0049/192/1/9

    2011 doi

  36. [44]

    E., et al.\ 2020, Nature Methods, 17, 261

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

    2020 doi

  37. [45]

    & Lewis, B

    Wurster, J. & Lewis, B. T.\ 2020, , 495, 3795. doi:10.1093/mnras/staa1339

    2020 doi

  38. [46]

    doi:10.1093/mnras/sty1165

    Zhao, B., Caselli, P., & Li, Z.-Y.\ 2018, , 478, 2723. doi:10.1093/mnras/sty1165

    2018 doi

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.