Pith. sign in

REVIEW 4 major objections 4 minor 38 references

X-ray Spectra from General Relativistic RMHD Simulations of Thin Disks

T0 review · 4 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read A coarse inline radiation scheme and a full Monte Carlo code agree on X-ray spectra from thin black hole disks, and the spectra harden monotonically with spin.

desk verdict A credible M1-vs-Monte Carlo transport comparison whose thermal-band agreement holds, but whose spin-hardening and full-spectrum claims lean too heavily on an unrelaxed middle-spin model. read the letter →

arxiv 2501.18040 v2 pith:FOGTHF2H submitted 2025-01-29 astro-ph.HE

classification astro-ph.HE
keywords blackholeaccretiondisksX-rayspectraradiationmagnetohydrodynamicsM1closureMonteCarloradiativetransferComptonscatteringspinbinaries
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 a relatively cheap, inline multi-frequency radiation-hydrodynamics treatment of thin black hole accretion disks produces X-ray spectra that agree with a much more detailed post-processing calculation. The answer it argues for is yes: across four simulations spanning accretion rates and black hole spins, the inline M1 method and the Monte Carlo transport code both yield a thermal peak near 2-2.5 keV, weaker thermally sourced emission out to about 100 keV, and an inverse-Compton component that grows with spin. The paper also claims a systematic, monotonic hardening of the spectrum with increasing black hole spin, with the highest-spin case reaching a photon index typical of the intermediate/hard state. If this is right, multi-frequency M1 simulations can be used directly to predict X-ray spectral shapes and spin trends without waiting for expensive full transport or relying on grey-opacity post-processing.

What carries the argument

The load-bearing comparison is between two radiation transport treatments applied to the same four disk states: the inline M1 two-moment closure, which tracks radiation energy density and net flux in twelve energy bins and treats Compton scattering through the Kompaneets equation, and a Monte Carlo transport code that follows photon bundles geodesically through the spacetime, scattering them with the Klein-Nishina cross section and a Maxwell-Juttner electron distribution. The M1 method's Kompaneets approximation is a second-order differential operator for photon diffusion in energy space, a low-energy Taylor expansion valid below the relativistic regime. The Monte Carlo method carries the argument because it resolves escaping radiation directionally and spectrally, and its birth-zone source maps identify where escaping photons originate, separating photosphere, inner-disk, and funnel emissions.

What would settle it

Take one spinning model such as S3Ea9 and continue the M1 simulation until the mass accretion rate settles out to at least 100 gravitational radii and the coronal temperature profile is steady over several ISCO periods, then repeat the Monte Carlo post-processing; if the 10-100 keV flux or the M1-MC agreement changes by more than the quoted variances, the spectral hardening is a stopping-time artifact rather than a spin property.

Watch

Extended reading notes

Core claim

The central discovery is that the M1 closure approximation, despite using only twelve logarithmically spaced energy bins and the Kompaneets treatment of Compton scattering, produces X-ray spectra that match those from a general-relativistic Monte Carlo code with fifty energy bins and full Klein-Nishina scattering. Both methods place the dominant soft-state thermal peak at 2-2.5 keV, agree on the weaker thermally sourced emission up to ~100 keV from the hot inner disk, and show inverse Compton scattering becoming the dominant emission mechanism between roughly 10 and several hundred keV. The spectrum hardens monotonically with spin: $a_*=0.75$ shows obvious Compton hardening, while $a_*=0.9$ produces a 5-100 keV photon index of $\Gamma=2$ and an upscattered flux near 100 keV comparable to the thermal flux, resembling the intermediate state. A fourth spectral component, free-free emission above $10^4$ keV from the hot, optically thin funnel near the horizon, appears in both spinning and non-spinning models. The authors attribute the spin trend to the inner disk edge moving closer to the black hole, producing hotter coronae that upscatter more photons.

Load-bearing premise

The simulations were stopped before the disks fully relaxed, so the spectra, especially the high-energy Compton component, may describe transient states rather than the settled disks they are meant to represent.

Editorial extensions

If this is right

  • If the M1-MC agreement holds, multi-frequency M1 simulations with a dozen energy bins can be used directly to predict X-ray spectral shapes of thin-disk black hole binaries without Monte Carlo post-processing for the 0.1-100 keV band.
  • The monotonic spin hardening provides a concrete, simulation-based connection between black hole spin and X-ray photon index: higher spin should push spectra from soft (index above 3.5) toward intermediate/hard (index near 2).
  • The four-component spectral decomposition (photosphere thermal peak, inner-disk thermal tail, inverse-Compton hump, coronal free-free) gives a physical map for interpreting observed black hole X-ray binary spectra in soft and intermediate states.
  • The highest-spin model's upscattered component approaching the thermal flux implies that intermediate-state-like spectra can be produced by spin alone, without invoking a separate Comptonizing corona.
  • The source maps show that photons escaping above $10^4$ keV are born in the optically thin funnel within a few gravitational radii of the horizon, making this emission a probe of the innermost accretion flow in both spinning and non-spinning cases.

Reading between the lines

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

  • A testable extension the authors do not pursue: run the same comparisons on simulations evolved past global relaxation, to separate the robust spin-hardening trend from transient coronal emission that may dominate the current S3Ea75 result.
  • If the M1 method is as reliable as claimed, its time-resolved fluxes could be used to generate spectral-timing predictions, such as lag-energy spectra, which the Monte Carlo post-processing approach is not optimized to produce.
  • The birth-zone maps suggest a way to test the disk-corona geometry directly: comparing the observed X-ray spectrum of a known black hole binary to the four-component decomposition could place constraints on the coronal temperature and the radius of the funnel emission, quantities the simulations currently fix.
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

4 major / 4 minor

Summary. The paper compares X-ray spectra from four general-relativistic, multi-frequency RMHD simulations of thin accretion disks, obtained with the inline M1 closure transport method, against spectra computed by post-processing the same hydrodynamic fields with a general-relativistic Monte Carlo transport code. The four models comprise two zero-spin cases at different accretion rates (S01Ea0, S3Ea0) and two spinning cases at nearly the same accretion rate (S3Ea75 at a*=0.75, S3Ea9 at a*=0.9). The manuscript describes the mapping of the spherical-polar simulation data onto Cartesian grids, the Monte Carlo setup, an idealized Kompaneets-vs-Monte-Carlo test in Section 3.2, and then presents spectral comparisons for the zero-spin and spinning models, source maps of escaping photon bundles, and a limited comparison with observations of black-hole X-ray binaries. The central claims are that M1 and MC produce generally consistent spectra, with strong agreement in the soft thermal peak near 2-2.5 keV and weaker thermally sourced emission extending to ~100 keV, and that increasing black-hole spin produces a systematic, monotonic spectral hardening driven by inverse Compton scattering. The paper is candid that the simulations are not fully relaxed, especially S3Ea75, which was run to only about 6,400 t_G.

Significance. If the central comparison holds, this is a useful validation of the coarse, inline multi-frequency M1 method against a higher-fidelity Monte Carlo transport calculation for thin-disk spectra, and the paper also demonstrates the diagnostic value of Monte Carlo post-processing through emission maps. The work is not circular: no parameters are fitted to force agreement, and the Monte Carlo solve is an independent transport calculation on the M1 hydrodynamic fields. The idealized Kompaneets test in Section 3.2 and the convergence checks on grid resolution, photon-bundle number, and time-dump choice are genuine strengths. However, the significance is limited by the unrelaxed nature of the simulations: the paper's own statements show that S3Ea75 is a highly dynamic transient state, and this model is used as the intermediate spin point in the monotonic-hardening claim and as a key case for the full-spectrum M1-MC agreement. The central claims therefore rest partly on the weakest data in the paper, and the conclusions currently overreach the evidence.

major comments (4)
  1. [§4.2, Table 1; §7] The monotonic spin-hardening claim is load-bearing, but its intermediate spin point S3Ea75 (a*=0.75) is the least reliable model. Table 1 reports t_stop=6,369 t_G versus >10,000 t_G for the other three models, and Section 4.2 states that S3Ea75 'remains in a highly dynamic transient state' and 'is not representative of steady-state emissions.' The same section notes that its high-energy spectrum is dominated by MC shot noise and by M1 flux-direction switching, including a drop at 10^-2 keV attributed to transient flux reversals. With S3Ea75 removed, the 'systematic (monotonic) hardening' claim reduces to two quasi-steady points, a*=0 and a*=0.9, which is not sufficient to establish monotonicity. The paper should either extend or relax S3Ea75, or revise the conclusion to state explicitly that the a*=0.75 point is provisional and that the hardening trend is established only between the two relaxed spin extremes.
  2. [§4.2, Figure 4] The full-spectrum M1-MC agreement claim is also carried by S3Ea75, yet Section 4.2 says that the exact spectral shape differs most 'particularly for the S3Ea75 calculation.' This undercuts the Conclusion's stronger statement that the two methods 'agree nicely across the entire spectrum.' In addition, the high-energy region of S3Ea75 is described as having a higher degree of Monte Carlo noise, so the agreement there is largely unquantified. The manuscript should either provide a quantitative agreement metric per model and energy band, or restrict the agreement claim to the thermal peak and to the qualitative shape of the high-energy tail, explicitly excluding the unrelaxed S3Ea75 spectrum from the full-spectrum claim.
  3. [§4.1, Figure 3, column (b)] For S3Ea0, the paper reports a visible disagreement between M1 and MC in the 30-100 keV band, where the M1 flux exceeds the MC flux. The text attributes this to relaxation behavior, spectral binning, or the idealized nature of the Kompaneets test, but does not quantify the discrepancy or demonstrate that it is within the stated uncertainties. Since this is one of the four models used to support the central method-comparison claim, the paper should either explain the discrepancy quantitatively with a comparison of temporal variances and statistical variances, or soften the 'generally good agreement' conclusion to specify that it applies primarily to the thermal peak and to the overall luminosity, with larger uncertainties in the intermediate-energy tail.
  4. [§4.2, Figure 4, column (b)] The paper shows that numerical energy corrections in the low-density funnel can artificially harden spectra, and it presents a masking test for S3Ea9 in the middle panel of column (b) of Figure 4. No analogous masking test is shown for S3Ea75 or for the zero-spin models. Because the spin-hardening claim depends on comparing S3Ea0, S3Ea75, and S3Ea9, the absence of a masking test for S3Ea75 is a load-bearing gap: the a*=0.75 spectrum could be affected by the same numerical artifact. Please add the corresponding masked spectra for S3Ea75, or justify why the funnel correction is negligible for that model.
minor comments (4)
  1. [Abstract and §7] The abstract and conclusions describe 'mass accretion rates (0.5 to 3c^2/L_Edd)', but Table 1 lists initial Novikov-Thorne rates of 0.01 and 3 and time-averaged rates of 0.52 and 0.66. Please specify which quantity is meant, or use consistent notation.
  2. [§4.1] There is a typo: 're-abosrbed' should be 're-absorbed'.
  3. [§3, mapping procedure] The paper maps the spherical-polar M1 data onto Cartesian grids using a nearest-neighbor, first-order procedure with azimuthal rotation, but does not demonstrate that this mapping preserves the spectral quantities of interest (especially the high-energy coronal emission). A short convergence test comparing spectra obtained with different mapping resolutions, or with the azimuthal rotation turned off, would increase confidence in the MC results.
  4. [§4.2] The sentence 'Although S3Ea75 is not representative of steady-state emissions, it remains a useful benchmark comparing transport methods, in addition to demonstrating the gradual hardening of the spectrum with increasing spin' is internally tension-inducing: if the model is not representative of steady state, its ability to demonstrate a monotonic physical trend is weakened. Please rephrase to clarify what exactly S3Ea75 is being used to demonstrate.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the M1-MC spectral comparison is an independent transport cross-check on shared hydrodynamic fields, not a fit or definitional reduction.

full rationale

The paper's central claims are comparisons between two independent radiation-transport solvers, not derivations that reduce to their inputs by construction. M1 spectra are reconstructed from evolved radiation fluxes via Eqs. (2)-(3), while MC spectra are tallied from photon bundles transported through the M1 hydrodynamic fields; the MC calculation does not read or reuse the M1 radiation field, so agreement between the methods is not enforced by construction. A wrong M1 Compton treatment or flux reconstruction would produce a mismatch, and indeed Section 4 reports energy-dependent disagreements (e.g., S3Ea0 between 30 and 100 keV, and the S3Ea75 spectral shape). No parameter is fitted to observed spectra, and the comparisons to LMC X-3 and Cyg X-1 are external benchmarks. The self-citations (Anninos & Fragile 2020; Fragile et al. 2023; Roth et al. 2022) are method references whose content is described in the text, including the Kompaneets equation (Eq. 4) and the MC scattering algorithm; they are not used as unverified uniqueness theorems or as hidden ansatz justifications. The acknowledged limitation that S3Ea75 is unrelaxed ('remains in a highly dynamic transient state') weakens the spin-hardening evidence, but that is a data-quality and validity concern, not a circular reduction. The MC spectra do inherit the M1 thermodynamics, so the soft thermal peak is partly a consistency check rather than a fully independent prediction, but this shared-input caveat does not make any claimed result equivalent to its input by definition.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities, forces, or particles. Its free parameters are numerical resolution, statistics, binning, and truncation choices, plus the funnel stability correction inherited from the M1 scheme. The most important assumptions are that the M1 closure and Kompaneets approximation are adequate, that the truncated simulations are representative, and that double Compton scattering can be neglected for the highest-energy claims.

free parameters (6)
  • MC grid resolution = 192^3 cells; min cell widths (0.06, 0.06, 0.0125) rG
    Chosen after convergence studies; coronal high-energy emission depends on resolution off the midplane (Section 3).
  • MC photon bundle budget = 5 bundles per zone per cycle, 20 cycles, about 2e9 total bundles
    Chosen as a compromise for statistics; high-energy bins above 100 keV are shot-noise limited (Sections 3 and 4.3).
  • Energy bin structure = M1: 12 bins from 5e-3 to 5e3 keV; MC: 50 bins from 1 to 1e8 eV
    The coarse M1 binning is part of the motivation for the comparison; bin ranges and boundaries are hand-chosen (Section 3.1).
  • Viewing cone and isotropization = 54 degree half-angle cone projected to 4pi
    Spectra are reported as isotropic-equivalent luminosities from a polar cone; this choice affects absolute luminosity values (Sections 3.1 and 4.1).
  • Funnel stability correction threshold = Magnetic pressure > 24 times gas pressure triggers temperature corrections
    Used in the original M1 runs to avoid numerical instabilities; the authors test sensitivity by excluding these cells in MC spectra, but the M1 output is still affected (Section 4.2).
  • Interaction and inner-radius cutoffs = Maximum 5e4 interactions per bundle; photons dropped within one cell width of the horizon
    Practical truncations to save compute time; the authors state there is no visible spectral effect but provide no quantitative demonstration (Section 3).
assumptions (5)
  • domain assumption The M1 two-moment closure with the Kompaneets approximation for Compton scattering is sufficiently accurate in the simulated disk and corona regimes.
    The paper's central comparison tests this axiom; Section 3.2 validates it only on a single idealized spherical corona model, not across all density and temperature regimes found in the disks.
  • domain assumption The hydrodynamic states mapped from the M1 runs at the final three data dumps are representative quasi-steady disk configurations.
    Section 2 admits that none of the runs fully relaxed and that S3Ea75 remains in a highly dynamic transient state at its stopping time. The spectral outputs depend on this representativeness assumption.
  • domain assumption Photon transport in these systems is dominated by free-free emission and absorption, Compton scattering, gravitational redshift, and Doppler shift; double Compton and stimulated Compton can be neglected.
    Section 3 states that stimulated and double Compton are not included, and Section 4.2 concedes that double Compton can dominate free-free emission from very hot disk regions. This assumption is load-bearing for the >10^4 keV free-free claim.
  • ad hoc to paper Nearest-neighbor mapping from spherical-polar to Cartesian grids, with azimuthal rotation, preserves the emitting structures relevant to the spectra.
    Section 3 describes a first-order nearest-neighbor mapping with no error estimate for the regridding step. The comparison between M1 and MC depends on this mapping being faithful.
  • standard math Kerr spacetime geodesic transport in the Monte Carlo code is accurate within the truncated inner boundary at one cell width from the horizon.
    Section 3 drops photons near the horizon to avoid tetrad normalization errors; the authors state that these rare photons do not visibly affect the spectra, but this is asserted rather than demonstrated quantitatively.

how reviews work

0 comments
Cite this review

Pith. "Pith review of X-ray Spectra from General Relativistic RMHD Simulations of Thin Disks." pith.science (2026). https://pith.science/paper/FOGTHF2H

@misc{pith2026250118040,
  author       = {Pith},
  title        = {Pith review of: X-ray Spectra from General Relativistic RMHD Simulations of Thin Disks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FOGTHF2H}},
  note         = {Machine review of arXiv:2501.18040}
}
abstract

We compare X-ray emission from several general relativistic, multi-frequency, radiation magnetohydrodynamic simulations of thin black hole accretion disks with different accretion rates and spins. The simulations were performed using the M1 closure scheme, resolved with twelve frequency (energy) bins logarithmically spaced from $5 \times 10^{-3}$ to $5 \times 10^3$ keV. We apply a general relativistic Monte Carlo transport code to post-process the simulation data with greater fidelity in frequency resolution and Compton scattering treatment. Despite the relatively few energy bins and Kompaneets approximation to Compton scattering utilized in the M1 method, we find generally good agreement between the methods. Both produce prominent thermal profiles with peaks around 2 - 2.5 keV, where agreement is particularly strong and representative of the soft state. Both also find weaker (lower luminosity) thermally sourced emission extending out to 100 keV due to the hotter innermost regions of the disks. Inverse Compton scattering becomes increasingly effective at hardening spectral outputs with increasing black hole spin, and becomes the dominant mechanism for photons that escape with energies between 10 to several hundred keV. At very high rates of spin the radiation flux in this upscattered component becomes comparable to the thermal flux, a phenomenon typically associated with intermediate states. Beyond $10^4$ keV, we observe faint, free-free emission from hot, optically thin coronal regions developing near the horizon, common to both spinning and nonspinning black holes.

Figures

Figures reproduced from arXiv: 2501.18040 by the authors.

Figure 1
Figure 1. Logarithms of the gas density (left) and temperature (right) mapped onto a geometrically refined Cartesian grid for models S3Ea0 (top) and S3Ea9 (bottom). The density is in units of g/cm3 , temperature in Kelvin, and the data correspond to a single snapshot (not time averaged) taken at the last simulation times. The color scales are normalized to the same minima and maxima in both models. The grid is designed to inc… view at source ↗
Figure 2
Figure 2. Spectral luminosities from two spherical models testing the accuracy of the Kompaneets approximation in resolving inverse Comptonization through hot coronal regions. The top (bottom) plot shows the spectrum with a corona temperature of 108 K (5 × 108 K), producing Compton peaks corresponding roughly to what we observe in our disk models. In both images, the Kompaneets solution is shown in red and the full MC transpo… view at source ↗
Figure 3
Figure 3. Spectral outputs derived from inlined M1 (red) and post-processed MC (black) calculations. Column (a) corre￾sponds to the S01Ea0 disk model, while column (b) corresponds to the S3Ea0 model. The top plot of each column includes Compton scattering, while the bottom plot replaces Compton scattering with Thomson. In all panels, the spectra represent the effective 4π outputs estimated from the flux traveling through a 54… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Similar to [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Statistical variance plots for cases S3Ea0 (top) and S3Ea9 (bottom), showing the Monte Carlo spectra along with 1σ contours representing accumulated shot noise. The spectra in both cases are computed from a total of more than 107 bundles across the entire energy grid. …
Figure 6
Figure 6. Figure 6: Spatial map plotting the logarithm of the average emission energy per volume (ergs/cm3 ) in the birth zones of escaping MC bundles, with energies between 0.1 - 1 keV (top row), 102 - 103 keV (second row), 103 - 104 keV (third row), and 104 - 105 keV (bottom). The maps …
Figure 7
Figure 7. Figure 7: Comparison of the post-processed Monte Carlo spectra from all four cases zoomed in to highlight the similarities in the dominant thermal peak at ∼ 2.5 keV and the differences across the inverse Compton scattered emissions between 10 and 103 keV. The spinning (non-spinn…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 7 canonical work pages

  1. [1]

    A., & Fragile, P

    Abramowicz, M. A., & Fragile, P. C. 2013, Living Reviews in Relativity, 16, 1, doi: 10.12942/lrr-2013-1

  2. [2]

    C., et al

    Anninos, P., Bryant, C., Fragile, P. C., et al. 2017, ApJS, 231, 17, doi: 10.3847/1538-4365/aa7ff5

  3. [3]

    Anninos, P., & Fragile, P. C. 2020, ApJ, 900, 71, doi: 10.3847/1538-4357/abab9c

  4. [4]

    C., & Salmonson, J

    Anninos, P., Fragile, P. C., & Salmonson, J. D. 2005, ApJ, 635, 723, doi: 10.1086/497294

  5. [5]

    F., & Krolik, J

    Beckwith, K., Hawley, J. F., & Krolik, J. H. 2008, ApJ, 678, 1180, doi: 10.1086/533492

  6. [6]

    C.-K., Ng, H

    Cheong, P. C.-K., Ng, H. H.-Y., Lam, A. T.-L., & Li, T. G. F. 2023, ApJS, 267, 38, doi: 10.3847/1538-4365/acd931

  7. [7]

    1971, PhRvD, 3, 2312, doi: 10.1103/PhysRevD.3.2312

    Cooper, G. 1971, PhRvD, 3, 2312, doi: 10.1103/PhysRevD.3.2312

  8. [8]

    W., Done, C., & Blaes, O

    Davis, S. W., Done, C., & Blaes, O. M. 2006, ApJ, 647, 525, doi: 10.1086/505386

Show all 38 references
  1. [9]

    2009, ApJ, 696, 1616, doi: 10.1088/0004-637X/696/2/1616

    Dexter, J., & Agol, E. 2009, ApJ, 696, 1616, doi: 10.1088/0004-637X/696/2/1616

  2. [10]

    Leung, P. K. 2009, ApJS, 184, 387, doi: 10.1088/0067-0049/184/2/387

  3. [11]

    2007, A&A Rv, 15, 1, doi: 10.1007/s00159-007-0006-1

    Done, C., Gierli´ nski, M., & Kubota, A. 2007, A&A Rv, 15, 1, doi: 10.1007/s00159-007-0006-1

  4. [12]

    1997, ApJL, 485, L87, doi: 10.1086/310816 Event Horizon Telescope Collaboration, Akiyama, K.,

    Dotani, T., Inoue, H., Mitsuda, K., et al. 1997, ApJL, 485, L87, doi: 10.1086/310816 Event Horizon Telescope Collaboration, Akiyama, K.,

  5. [13]

    2019, ApJL, 875, L5, doi: 10.3847/2041-8213/ab0f43 —

    Alberdi, A., et al. 2019, ApJL, 875, L5, doi: 10.3847/2041-8213/ab0f43 —. 2022, ApJL, 930, L16, doi: 10.3847/2041-8213/ac6672

  6. [14]

    C., Anninos, P., Roth, N., & Mishra, B

    Fragile, P. C., Anninos, P., Roth, N., & Mishra, B. 2023, ApJ, 959, 59, doi: 10.3847/1538-4357/ad096b

  7. [15]

    C., Nemergut, D., Shaw, P

    Fragile, P. C., Nemergut, D., Shaw, P. L., & Anninos, P. 2019, Journal of Computational Physics: X, 2, 100020, doi: https://doi.org/10.1016/j.jcpx.2019.100020

  8. [16]

    C., Olejar, A., & Anninos, P

    Fragile, P. C., Olejar, A., & Anninos, P. 2014, ApJ, 796, 22, doi: 10.1088/0004-637X/796/1/22 Gierli´ nski, M., & Done, C. 2003, MNRAS, 342, 1083, doi: 10.1046/j.1365-8711.2003.06591.x —. 2004, MNRAS, 347, 885, doi: 10.1111/j.1365-2966.2004.07266.x

  9. [17]

    2022, ApJS, 263, 4, doi: 10.3847/1538-4365/ac9231

    Jiang, Y.-F. 2022, ApJS, 263, 4, doi: 10.3847/1538-4365/ac9231

  10. [18]

    M., & Davis, S

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

  11. [19]

    Kawashima, T., Ohsuga, K., & Takahashi, H. R. 2023, ApJ, 949, 101, doi: 10.3847/1538-4357/acc94a

  12. [20]

    E., Schnittman, J

    Kinch, B. E., Schnittman, J. D., Kallman, T. R., & Krolik, J. H. 2019, ApJ, 873, 71, doi: 10.3847/1538-4357/ab05d5

  13. [21]

    E., Schnittman, J

    Kinch, B. E., Schnittman, J. D., Noble, S. C., Kallman, T. R., & Krolik, J. H. 2021, ApJ, 922, 270, doi: 10.3847/1538-4357/ac2b9a

  14. [22]

    Kompaneets, A. S. 1957, Soviet Journal of Experimental and Theoretical Physics, 4, 730

  15. [23]

    E., & Remillard, R

    McClintock, J. E., & Remillard, R. A. 2006, in Compact stellar X-ray sources, ed. W. H. G. Lewin & M. van der

  16. [24]

    39, 157–213, doi: 10.48550/arXiv.astro-ph/0306213

    Klis, Vol. 39, 157–213, doi: 10.48550/arXiv.astro-ph/0306213

  17. [25]

    2017, MNRAS, 467, 2241, doi: 10.1093/mnras/stx227

    Sadowski, A. 2017, MNRAS, 467, 2241, doi: 10.1093/mnras/stx227

  18. [26]

    2014, MNRAS, 441, 3177, doi: 10.1093/mnras/stu762

    Narayan, R. 2014, MNRAS, 441, 3177, doi: 10.1093/mnras/stu762

  19. [27]

    S., Davis, S

    Mills, B. S., Davis, S. W., Jiang, Y.-F., & Middleton, M. J. 2024, ApJ, 974, 166, doi: 10.3847/1538-4357/ad6b21

  20. [28]

    2016, MNRAS, 457, 608, doi: 10.1093/mnras/stv2979

    Narayan, R., Zhu, Y., Psaltis, D., & S¸ adowski, A. 2016, MNRAS, 457, 608, doi: 10.1093/mnras/stv2979

  21. [29]

    D., & Thorne, K

    Novikov, I. D., & Thorne, K. S. 1973, in Black Holes (Les Astres Occlus), 343–450

  22. [30]

    B., et al

    Roth, N., Anninos, P., Robinson, P. B., et al. 2022, ApJ, 933, 226, doi: 10.3847/1538-4357/ac75cb S¸ adowski, A., Narayan, R., McKinney, J. C., &

  23. [31]

    2014, MNRAS, 439, 503, doi: 10.1093/mnras/stt2479

    Tchekhovskoy, A. 2014, MNRAS, 439, 503, doi: 10.1093/mnras/stt2479

  24. [32]

    D., Krolik, J

    Schnittman, J. D., Krolik, J. H., & Noble, S. C. 2013, ApJ, 769, 156, doi: 10.1088/0004-637X/769/2/156

  25. [33]

    E., Narayan, R., et al

    Shafee, R., McClintock, J. E., Narayan, R., et al. 2006, ApJL, 636, L113, doi: 10.1086/498938

  26. [34]

    Tchekhovskoy, A., Narayan, R., & McKinney, J. C. 2010, ApJ, 711, 50, doi: 10.1088/0004-637X/711/1/50

  27. [35]

    Thorne, K. S. 1981, MNRAS, 194, 439, doi: 10.1093/mnras/194.2.439

  28. [36]

    Wilkins, D. R. 2014, A&A Rv, 22, 72, doi: 10.1007/s00159-014-0072-0

  29. [37]

    H., Paumard, T., Gourgoulhon, E., & Perrin, G

    Vincent, F. H., Paumard, T., Gourgoulhon, E., & Perrin, G. 2011, Classical and Quantum Gravity, 28, 225011, doi: 10.1088/0264-9381/28/22/225011 17

  30. [38]

    A., Poutanen, J., Paciesas, W

    Zdziarski, A. A., Poutanen, J., Paciesas, W. S., & Wen, L. 2002, ApJ, 578, 357, doi: 10.1086/342402

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.