Pith. sign in

REVIEW 4 major objections 5 minor 61 references

FEADME: Fast Elliptical Accretion Disk Modeling Engine

T0 review · 4 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Pith's one-line read Modeling 237 AGN and five tidal disruption events with a uniform relativistic elliptical-disk fit, this paper finds that TDE disks are significantly more circular than AGN disks (median eccentricity 0.17 versus 0.64), while otherwise overla

desk verdict A useful tool paper whose main astrophysical claim is undermined by epoch-level pseudo-replication and an untested degeneracy with the broad Gaussian component. read the letter →

arxiv 2512.10228 v2 pith:E4LEVKP7 submitted 2025-12-11 astro-ph.HE

classification astro-ph.HE
keywords accretiondiskstidaldisruptioneventsdouble-peakedemittersbroadBalmerlinesellipticaldiskmodeleccentricityBayesianinferenceAGN
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 builds a fast Bayesian fitting engine for relativistic elliptical accretion-disk line profiles and applies it uniformly to hundreds of AGN double-peaked emitters and five tidal disruption events. It claims that, once a line-emitting disk has formed, TDE and AGN disks are spectroscopically nearly indistinguishable: their radial scales, inclinations, emissivity slopes, and local broadening overlap, and most objects in both populations need an extra broad Gaussian component on top of the disk. The one robust difference is eccentricity: TDE disks come out rounder (median 0.17 versus 0.64) with high significance, which the authors read as evidence of rapid debris circularization in TDEs. If right, this means the physics of line formation in relativistic disks is universal across transient and persistent accretion flows, and that fitted disk eccentricity can trace the dynamical history of the accreting gas.

What carries the argument

The load-bearing object is the relativistic elliptical accretion-disk line-profile model: geometrically thin, optically thick Keplerian elliptical orbits around a black hole, with a power-law radial emissivity and local Gaussian broadening. Its seven parameters generate the double-peaked H-alpha profile through a Doppler-boosted integral over the disk, including light-bending corrections. The paper implements this model as a differentiable forward model and fits it with Hamiltonian Monte Carlo, comparing three model families (disk + broad Gaussian + narrow lines; disk only; broad Gaussian only) by approximate leave-one-out cross-validation. The fitted disk eccentricity, co-fit with the broad

What would settle it

Synthetic injection: generate mock double-peaked profiles with known disk eccentricity, fit them with the full model (disk + free broad Gaussian), and compare recovered versus input eccentricity as a function of broad Gaussian width. If broader Gaussians systematically lower recovered eccentricity, the TDE-versus-AGN gap is plausibly degenerate; if recovered eccentricity is robust to broad-Gaussian width, the population difference is physical.

Watch

Extended reading notes

Core claim

The central discovery is that the eccentricity of the line-emitting disk separates the two populations. Using the same seven-parameter relativistic thin-disk model (inner and outer radius, eccentricity, inclination, apocenter angle, emissivity index, local broadening) fit with Bayesian inference and model selection, the paper finds that TDE disks have median e = 0.17 versus an AGN median of e = 0.64, with a Mann–Whitney p-value of 2.3e-8. For inclination, emissivity slope, local broadening, and outer radius, the TDE and AGN distributions are statistically indistinguishable (p > 0.1 for five of seven parameters); inner radius is smaller in TDEs but within the AGN range. The non-disk broad Gau

Load-bearing premise

The fitted disk eccentricity is a physical orbital property rather than a trade-off with the co-fit broad Gaussian component; the paper does not test this degeneracy, so the reported roundness of TDE disks could be an artifact of model flexibility.

Editorial extensions

If this is right

  • If TDE disks truly are rounder, then the emergence of broad Balmer emission marks a state in which stellar debris has already circularized; low eccentricity can serve as a spectroscopic clock for disk formation in future TDEs.
  • The near-universal coexistence of disk and extra broad component implies that disk emission is rarely isolated; single-component models of broad lines will systematically misattribute flux.
  • The overlap in other parameters means that TDEs can be used as laboratories for AGN disk physics with time-resolved spectroscopy, since the same model applies to both.
  • The broader non-disk component in TDEs implies that the high-velocity gas in TDEs is dynamically younger and more energetic than the virialized broad-line region of AGN, a difference that should show up in multi-line comparisons.
  • Population-scale, reproducible disk modeling becomes feasible, allowing future surveys to fit hundreds of sources with the same code and physical assumptions.

Reading between the lines

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

  • The eccentricity difference may partly reflect a model degeneracy: the extra broad Gaussian, which is significantly wider in TDEs, could absorb asymmetric line wings that a narrower Gaussian would leave for the disk's eccentricity to explain. A direct test would be to fit AGN with TDE-like broad Gaussian widths and see whether inferred eccentricities drop toward 0.2.
  • Because the TDE sample was preselected for double-peaked morphology and is tiny (five events), the overlap with AGN may be a selection effect; including TDEs without double-peaked profiles would test whether the round-disk conclusion extends to all TDE disks.
  • If the thin-disk approximation is violated for super-Eddington TDEs, the fitted 'eccentricity' may be a proxy for the asymmetry of a thickened disk or wind; then the correct physical statement would be that TDE line-forming regions are more symmetric, not more circular—a distinction testable with radiative-transfer models of thick disks and winds.
  • The time-resolved fits show early epochs with elevated eccentricity in at least two events, so an explicit testable prediction is that eccentricity decreases monotonically with time in individual TDEs as circularization proceeds; the current sample is too sparse to confirm this.
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 / 5 minor

Summary. The paper presents FEADME, a GPU-accelerated JAX/NumPyro implementation of the Eracleous et al. (1995) relativistic elliptical disk model for broad Balmer-line profiles. The authors fit three model families (disk + narrow lines + broad Gaussian, disk + narrow lines only, and broad Gaussian + narrow lines only) to 237 AGN double-peaked emitters from Ward et al. (2024) and to five TDEs, select a preferred model per spectrum with LOO/WAIC, and compare posterior parameter distributions. The main astrophysical claim is that TDE disks are significantly less eccentric (median e = 0.17) than AGN disks (median e = 0.64; p = 2.3e-8), while other disk parameters overlap; a secondary claim is that the non-disk broad Gaussian component is broader in TDEs.

Significance. If the eccentricity difference is robust, it is an interesting and timely constraint on debris circularization and it supports a unified picture of line-forming disks in persistent AGN and transient TDEs. The paper also contributes an open-source, reproducible modeling framework and applies a consistent Bayesian pipeline to a large sample, which are genuine strengths. However, the headline result rests on two currently untested statistical assumptions: no degeneracy between disk eccentricity and the broad Gaussian component, and independence of repeated TDE epochs. The manuscript also contains internal inconsistencies in sample sizes and model-selection terminology. The central claim is therefore not yet established; targeted revisions could make it publishable.

major comments (4)
  1. [§6.1, Table 2, Fig. 7] The conclusion that TDE disks are rounder is potentially confounded by the simultaneously broader non-disk Gaussian component. TDEs require a broad Gaussian with median FWHM ≈ 1.1e4 km/s versus ≈ 6.7e3 km/s for AGN (p = 4e-5). The paper never tests whether a broad Gaussian can absorb the line asymmetries that would otherwise drive e upward. Please report posterior correlations between e and Gaussian FWHM, include a circular-disk + broad-Gaussian model in the TDE model comparison, or run simulations with known e to show that the fitted e is not biased low by the extra component.
  2. [§6.1, Table 2] The Mann–Whitney test compares 27 (or 29) TDE epochs drawn from only five events against the AGN source distribution. Repeated epochs of the same transient trace the same disk and viewing geometry and are serially correlated; pooling them as independent samples inflates the effective sample size by roughly a factor of five. The p = 2.3e-8 therefore does not establish that 'TDE disks are rounder' as a population statement. The authors should report an event-level comparison (e.g., one summary e per TDE, or a hierarchical model) and recompute the significance.
  3. [Abstract, §4, §5.2, Table 1] Sample-size and model-selection statements are internally inconsistent. The abstract says 237 AGN and 165 AGN after filtering, 27 usable TDE epochs, and WAIC; §4 says 29 usable TDE epochs and uses LOO; §5.2 clustering sums to 199 AGN (167 + 32) rather than 237; and the abstract's 'five phenomenological Gaussian-mixture morphology bins' does not match the four HDBSCAN clusters described in the text. Please reconcile all numbers, define the posterior-quality filtering, and use one model-selection terminology throughout.
  4. [§6.1, §7] The statement that 'with the exception of eccentricity, TDE disks are not strongly distinguishable from AGN disks' is contradicted by the paper's own tests: the broad Gaussian FWHM differs at p = 4.18e-5 and the inner radius at p = 2.1e-3. These are reported in §6.1 as significant differences. Please either qualify the conclusion to include these parameters or explicitly explain why these differences are not considered 'strong' in the context of the paper's claims.
minor comments (5)
  1. [§2.2] Typo: 'ASSASN-14li' should be 'ASASSN-14li'.
  2. [§6.2] Typo: 'standard think-disk solutions' should read 'standard thin-disk solutions'; also 'Eddington ratio of ∼0.6at peak' is missing a space.
  3. [§4] The model-family label 'No-BLR Model' is confusing because the 'BLR' component in the Full Model is a broad Gaussian, not the full broad-line region; consider renaming to 'No-Broad-Gaussian'.
  4. [§5.2] The clustering analysis reports four clusters, but Table 1 and Figure 5 do not list a fifth cluster; the abstract's 'five ... morphology bins' appears to be a residual from an earlier draft and should be corrected.
  5. [Abstract] The abstract uses 'widely applicable information criterion (WAIC)' while the text uses approximate leave-one-out (LOO) cross-validation. Choose one and use it consistently.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central result is an empirical posterior comparison, not a derivation from fitted inputs.

full rationale

The paper does not claim to derive TDE eccentricity from first principles. All disk parameters (e, i, ξ1, ξ2, q, σ, φ0) are free parameters in the adopted Eracleous et al. (1995) elliptical disk model, fitted independently to each spectrum. The headline result—TDE median e = 0.17 versus AGN median e = 0.64, p = 2.3e-8—is a Mann-Whitney comparison of posterior medians, so there is no fitted-input-called-prediction reduction: the 'prediction' is the posterior itself, not an independent quantity derived from it. The model is adopted from an external, standard reference (Eracleous et al. 1995; Strateva et al. 2003), not from a self-citation chain, and FEADME is an implementation rather than a derivation. The only self-citations (e.g., N. Earl et al. 2025 for AT2020nov's prior outer radius and eccentricity) are contextual; Table 2 reports the paper's own refits, so the central eccentricity comparison does not depend on those cited values. The acknowledged thin-disk limitation in §6.2 is a caveat about interpretation, not a circular step. Statistical concerns such as epoch-level pseudo-replication and possible degeneracy between e and the broad Gaussian are correctness and robustness risks, not circularity by construction.

Assumptions & free parameters 8 free parameters · 7 assumptions · 0 invented entities

No new physical entities are introduced. The ledger lists the seven (plus broad-Gaussian) fitted parameters and the modeling assumptions inherited from Eracleous-style thin-disk models, plus the untested assumption that the broad Gaussian does not degenerate with eccentricity.

free parameters (8)
  • e (disk eccentricity) = AGN median 0.64; TDE median 0.17
    Central to the main claim; fitted per spectrum; the comparison of this parameter drives the paper's headline result.
  • i (inclination) = AGN median 61 deg; TDE median 64 deg
    Fitted disk orientation; needed to produce double-peaked profiles.
  • xi1 (inner semimajor axis) = AGN median 2225 Rg
    Fitted with a log-uniform prior 100-10000 Rg; part of the seven-parameter disk model.
  • xi2 (outer semimajor axis) = AGN median 11882 Rg
    Fitted via an undocumented 'radius_scale' parameter in the template rather than directly as xi2; the text's Eq. 8 lists xi2 as free.
  • q (emissivity index) = AGN median 1.69
    Power-law exponent for radial emissivity; fitted per spectrum.
  • sigma (local Gaussian broadening) = AGN median 819 km/s; TDE median 649 km/s
    Accounts for non-Keplerian motions; fitted per spectrum.
  • phi0 (apocenter angle) = AGN median 220 deg
    Azimuthal orientation of the elliptical disk; fitted per spectrum.
  • Broad Gaussian FWHM = TDE median ~1.1e4 km/s; AGN median ~6.7e3 km/s
    Width of the extra Gaussian component; potential confound with eccentricity, untested.
assumptions (7)
  • domain assumption The Eracleous et al. (1995) weak-field relativistic elliptical disk model (Eqs. 2-6) accurately describes the shape of the broad Balmer-line profile.
    Invoked throughout §3; the light-bending factor is a weak-field approximation and the model is adopted from prior literature without independent validation here.
  • domain assumption Line-emitting gas is geometrically thin, optically thick, coplanar, and follows Keplerian ellipses with a single common eccentricity e.
    §3.2 and §6.2; acknowledged as 'more tentative' for TDEs, where flows may be thick and wind-dominated.
  • domain assumption Emissivity follows xi^{-q} and local broadening is Gaussian with a single sigma.
    §3.2 Eq. 5; adopted from the Eracleous model, not derived here.
  • domain assumption Normalized profile shape (amplitude factored out) is sufficient, so unknown black-hole masses and distances do not affect the fit.
    §3.2; the model discards absolute flux information, relying on shape only.
  • ad hoc to paper The broad Gaussian component captures non-disk emission and does not strongly degenerate with disk eccentricity.
    Untested; central to the eccentricity comparison in §6.1 and flagged as a red flag.
  • domain assumption Leave-one-out cross-validation (ArviZ) reliably selects the correct model family for each spectrum.
    §4; standard technique, but its reliability for strongly structured spectral models with correlated parameters is not demonstrated.
  • domain assumption The five selected TDEs are representative of disk-like TDEs, and multi-epoch stacking weights epochs within an event equally.
    §2.2, §5.3; AT2020nov contributes about 12 epochs, biasing stacked comparisons toward that single event.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FEADME: Fast Elliptical Accretion Disk Modeling Engine." pith.science (2026). https://pith.science/paper/E4LEVKP7

@misc{pith2026251210228,
  author       = {Pith},
  title        = {Pith review of: FEADME: Fast Elliptical Accretion Disk Modeling Engine},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/E4LEVKP7}},
  note         = {Machine review of arXiv:2512.10228}
}
abstract

We present FEADME (Fast Elliptical Accretion Disk Modeling Engine), a GPU-accelerated Python framework for modeling broad Balmer-line emission using a relativistic elliptical accretion-disk formalism. Leveraging JAX and NumPyro for differentiable forward modeling and efficient Bayesian inference, FEADME enables large-sample, reproducible analyses of disk-dominated emission-line profiles. We apply the framework to 237 double-peaked emitters (DPEs) from the literature and to five tidal disruption events (TDEs) with disk-like H$\alpha$ emission, fitting three physically motivated model families per spectrum and selecting the preferred model using the widely applicable information criterion (WAIC). After posterior-quality filtering, the disk-bearing active galactic nuclei (AGN) analysis sample contains 165 sources and the TDE sample contains 27 usable epochs. We find that AGN occupy a broad, continuous distribution of disk geometries and kinematics that is usefully summarized by five phenomenological Gaussian-mixture morphology bins. The TDE disk parameters overlap substantially with the AGN population in radial scale, local broadening, and emissivity slope, but TDEs are systematically less eccentric and show broader non-disk Gaussian components. The majority of both AGN and TDEs favor models that include both a disk and an additional broad-line component, suggesting that disk emission commonly coexists with more isotropic or wind-driven gas. These results indicate that once a line-emitting disk forms, its spectroscopic appearance is governed by similar physical processes in both persistent AGN and transient TDE accretion flows, and they demonstrate the utility of FEADME for population-level studies of disk structure in galactic nuclei.

Figures

Figures reproduced from arXiv: 2512.10228 by the authors.

Figure 1
Figure 1. Representative rest-frame spectra for each of the five TDEs in our sample (AT 2018hyz, AT 2018zr, AT 2020nov, AT 2020zso, and PTF09djl) showing broad, double-peaked Hα emission consistent with disk-like kine￾matics. Each panel displays a single epoch selected to high￾light the double-peaked structure, with the offset from peak optical brightness shown in the top-left corner. The diversity in peak separation, asymmet… view at source ↗
Figure 2
Figure 2. Representative examples of the three model families used to fit the AGN spectra in this work. Each panel shows an observed AGN spectrum (grey), together with the corresponding best-fit model (black dashed line). The contribution from the elliptical accretion disk is shown in gold, while all non-disk line-emission components (narrow lines and the broad Gaussian) are shown in cyan. These examples illustrate the qualit… view at source ↗
Figure 3
Figure 3. Distribution of elliptical disk model parameters across AGN clusters identified via UMAP + HDBScan clustering (see Section 5.2), as well as for the full AGN sample (light grey) and the full TDE sample (dark grey). Each panel corresponds to one of the eight model parameters: the seven elliptical disk parameters (inclination (i), emissivity index (q), turbulent broadening (σ), inner radius (ξ1), outer radius (ξ2), apo… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Two-dimensional projection of the 5-compo￾nent UMAP embedding of the AGN disk parameter space. Colored regions indicate high-density areas associated with the four HDBScan-identified AGN clusters, while individ￾ual spectra not assigned to any cluster (i.e. transitional…
Figure 5
Figure 5. Figure 5: Median rest-frame Hα fitted disk profiles for each of the four AGN clusters identified via UMAP + HDBScan. Solid lines represent the median spectrum within each clus￾ter, while the shaded regions denote the 16th–84th percentile range (dark shading) and the 5th–95th per…
Figure 6
Figure 6. Figure 6: Time evolution of the fitted elliptical disk model parameters for the five TDEs in our sample. Each panel shows one of the eight parameters (the seven disk parameters along with the FWHM of the broad Gaussian component when included). Time is normalized individually fo…
Figure 7
Figure 7. Figure 7: Comparison of posterior parameter distributions between the AGN and TDE samples. Each panel shows one of the eight fitted parameters (the seven elliptical disk parameters, plus the FWHM of the broad Gaussian component, when included). Grey histograms represent the stac…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

61 extracted references · 10 canonical work pages

  1. [1]

    C., Kulkarni, S

    Bellm, E. C., Kulkarni, S. R., Graham, M. J., et al. 2019, PASP, 131, 018002, doi: 10.1088/1538-3873/aaecbe

  2. [2]

    P., Jankowiak, M., et al

    Bingham, E., Chen, J. P., Jankowiak, M., et al. 2018, arXiv e-prints, arXiv:1810.09538, doi: 10.48550/arXiv.1810.09538

  3. [4]

    2020, MNRAS, 495, 1374, doi: 10.1093/mnras/staa1246

    Bonnerot, C., & Lu, W. 2020, MNRAS, 495, 1374, doi: 10.1093/mnras/staa1246

  4. [5]

    Campello, R. J. G. B., Moulavi, D., & Sander, J. 2013, in Advances in Knowledge Discovery and Data Mining, ed. J. Pei, V. S. Tseng, L. Cao, H. Motoda, & G. Xu (Berlin, Heidelberg: Springer Berlin Heidelberg), 160–172

  5. [6]

    K., Zhou, Z

    Cao, R., Liu, F. K., Zhou, Z. Q., Komossa, S., & Ho, L. C. 2018, MNRAS, 480, 2929, doi: 10.1093/mnras/sty1997

  6. [7]

    2023, MNRAS, 526, 3273, doi: 10.1093/mnras/stad2597

    Cappellari, M. 2023, MNRAS, 526, 3273, doi: 10.1093/mnras/stad2597

  7. [8]

    S., & Hall, P

    Chajet, L. S., & Hall, P. B. 2013, MNRAS, 429, 3214, doi: 10.1093/mnras/sts580

  8. [9]

    B., et al

    Charalampopoulos, P., Leloudas, G., Malesani, D. B., et al. 2022, A&A, 659, A34, doi: 10.1051/0004-6361/202142122 19

Show all 61 references
  1. [10]

    P., & Filippenko, A

    Chen, K., Halpern, J. P., & Filippenko, A. V. 1989, ApJ, 339, 742, doi: 10.1086/167332

  2. [11]

    Miller, M. C. 2018, ApJL, 859, L20, doi: 10.3847/2041-8213/aab429

  3. [12]

    M., Riddle, R., et al

    Dekany, R., Smith, R. M., Riddle, R., et al. 2020, PASP, 132, 038001, doi: 10.1088/1538-3873/ab4ca2

  4. [13]

    D., Ramirez-Ruiz, E., et al

    Earl, N., French, K. D., Ramirez-Ruiz, E., et al. 2025, ApJ, 983, 28, doi: 10.3847/1538-4357/adb974

  5. [14]

    1998, Advances in Space Research, 21, 33, doi: 10.1016/S0273-1177(97)00612-1

    Eracleous, M. 1998, Advances in Space Research, 21, 33, doi: 10.1016/S0273-1177(97)00612-1

  6. [15]

    Eracleous, M., & Halpern, J. P. 2003, ApJ, 599, 886, doi: 10.1086/379540

  7. [16]

    1995, ApJ, 438, 610, doi: 10.1086/175104

    Storchi-Bergmann, T. 1995, ApJ, 438, 610, doi: 10.1086/175104

  8. [17]

    Flohic, H. M. L. G., Eracleous, M., & Bogdanović, T. 2012, ApJ, 753, 133, doi: 10.1088/0004-637X/753/2/133

  9. [18]

    P., & Eracleous, M

    Gezari, S., Halpern, J. P., & Eracleous, M. 2007, ApJS, 169, 167, doi: 10.1086/511032

  10. [19]

    2019, A&A, 630, A94, doi: 10.1051/0004-6361/201833810

    Giustini, M., & Proga, D. 2019, A&A, 630, A94, doi: 10.1051/0004-6361/201833810

  11. [20]

    J., Kulkarni, S

    Graham, M. J., Kulkarni, S. R., Bellm, E. C., et al. 2019, PASP, 131, 078001, doi: 10.1088/1538-3873/ab006c GRAVITY Collaboration, Amorim, A., Bourdarot, G., et al. 2024, A&A, 684, A167, doi: 10.1051/0004-6361/202348167

  12. [21]

    C., Rudnick, G., Rix, H.-W., et al

    Ho, L. C., Rudnick, G., Rix, H.-W., et al. 2000, ApJ, 541, 120, doi: 10.1086/309440

  13. [22]

    V., et al

    Hoffman, M., Sountsov, P., Dillon, J. V., et al. 2019, arXiv e-prints, arXiv:1903.03704, doi: 10.48550/arXiv.1903.03704

  14. [23]

    Holoien, T. W. S., Huber, M. E., Shappee, B. J., et al. 2019, ApJ, 880, 120, doi: 10.3847/1538-4357/ab2ae1

  15. [24]

    Hopkins, B., & Skellam, J. G. 1954, Annals of Botany, 18, 213, doi: 10.1093/oxfordjournals.aob.a083391

  16. [25]

    W., & Jiang, Y.-f

    Huang, X., Davis, S. W., & Jiang, Y.-f. 2024, ApJ, 974, 165, doi: 10.3847/1538-4357/ad6c39

  17. [26]

    J., Ramirez-Ruiz, E., et al

    Hung, T., Foley, R. J., Ramirez-Ruiz, E., et al. 2020, The Astrophysical Journal, 903, 31, doi: 10.3847/1538-4357/abb606

  18. [27]

    2012, A&A, 548, A37, doi: 10.1051/0004-6361/201118174

    Kollatschny, W., Reichstein, A., & Zetzl, M. 2012, A&A, 548, A37, doi: 10.1051/0004-6361/201118174

  19. [28]

    2013, A&A, 549, A100, doi: 10.1051/0004-6361/201219411

    Kollatschny, W., & Zetzl, M. 2013, A&A, 549, A100, doi: 10.1051/0004-6361/201219411

  20. [29]

    Krolik, J., Piran, T., Svirski, G., & Cheng, R. M. 2016, ApJ, 827, 127, doi: 10.3847/0004-637X/827/2/127

  21. [30]

    T., Eracleous, M., & Storchi-Bergmann, T

    Lewis, K. T., Eracleous, M., & Storchi-Bergmann, T. 2010, ApJS, 187, 416, doi: 10.1088/0067-0049/187/2/416

  22. [31]

    K., Zhou, Z

    Liu, F. K., Zhou, Z. Q., Cao, R., Ho, L. C., & Komossa, S. 2017, MNRAS, 472, L99, doi: 10.1093/mnrasl/slx147

  23. [32]

    2018, arXiv e-prints, arXiv:1802.03426, doi: 10.48550/arXiv.1802.03426

    McInnes, L., Healy, J., & Melville, J. 2018, arXiv e-prints, arXiv:1802.03426, doi: 10.48550/arXiv.1802.03426

  24. [33]

    C., Dai, L., & Avara, M

    McKinney, J. C., Dai, L., & Avara, M. J. 2015, MNRAS, 454, L6, doi: 10.1093/mnrasl/slv115

  25. [34]

    D., & Stone, N

    Metzger, B. D., & Stone, N. C. 2016, MNRAS, 461, 948, doi: 10.1093/mnras/stw1394

  26. [35]

    1996, Nature, 382, 789, doi: 10.1038/382789a0

    Murray, N., & Chiang, J. 1996, Nature, 382, 789, doi: 10.1038/382789a0

  27. [36]

    M., & Sánchez, F

    Negus, J., Comerford, J. M., & Sánchez, F. M. 2024, ApJ, 971, 92, doi: 10.3847/1538-4357/ad4c68

  28. [37]

    2011, ApJ, 736, 2, doi: 10.1088/0004-637X/736/1/2

    Ohsuga, K., & Mineshige, S. 2011, ApJ, 736, 2, doi: 10.1088/0004-637X/736/1/2

  29. [38]

    I., Blackman, E

    Pariev, V. I., Blackman, E. G., & Boldyrev, S. A. 2003, A&A, 407, 403, doi: 10.1051/0004-6361:20030868

  30. [39]

    J., Knigge, C., Matthews, J

    Parkinson, E. J., Knigge, C., Matthews, J. H., et al. 2022, MNRAS, 510, 5426, doi: 10.1093/mnras/stac027

  31. [40]

    2019, arXiv e-prints, arXiv:1912.11554, doi: 10.48550/arXiv.1912.11554

    Phan, D., Pradhan, N., & Jankowiak, M. 2019, arXiv e-prints, arXiv:1912.11554, doi: 10.48550/arXiv.1912.11554

  32. [41]

    2015, ApJ, 806, 164, doi: 10.1088/0004-637X/806/2/164 Popović, L

    Shiokawa, H. 2015, ApJ, 806, 164, doi: 10.1088/0004-637X/806/2/164 Popović, L. Č., Mediavilla, E., Bon, E., & Ilić, D. 2004, A&A, 423, 909, doi: 10.1051/0004-6361:20034431

  33. [42]

    M., & Kallman, T

    Proga, D., Stone, J. M., & Kallman, T. R. 2000, ApJ, 543, 686, doi: 10.1086/317154

  34. [43]

    Rees, M. J. 1988, Nature, 333, 523, doi: 10.1038/333523a0

  35. [44]

    2012, ApJ, 748, 145, doi: 10.1088/0004-637X/748/2/145

    Winge, C., & Eracleous, M. 2012, ApJ, 748, 145, doi: 10.1088/0004-637X/748/2/145

  36. [45]

    S., & Eracleous, M

    Nemmen, R. S., & Eracleous, M. 2017, MNRAS, 472, 2170, doi: 10.1093/mnras/stx2107

  37. [46]

    I., & Sunyaev, R

    Shakura, N. I., & Sunyaev, R. A. 1973, A&A, 24, 337

  38. [47]

    2009, New Astronomy Reviews, 53, 191, doi: https://doi.org/10.1016/j.newar.2009.08.004

    Shapovalova, A., Popović, L., Bochkarev, N., et al. 2009, New Astronomy Reviews, 53, 191, doi: https://doi.org/10.1016/j.newar.2009.08.004

  39. [48]

    Noble, S. C. 2015, ApJ, 804, 85, doi: 10.1088/0004-637X/804/2/85

  40. [49]

    2020, MNRAS, 498, 4119, doi: 10.1093/mnras/staa2065 Sądowski, A., Tejeda, E., Gafton, E., Rosswog, S., &

    Short, P., Nicholl, M., Lawrence, A., et al. 2020, MNRAS, 498, 4119, doi: 10.1093/mnras/staa2065 Sądowski, A., Tejeda, E., Gafton, E., Rosswog, S., &

  41. [50]

    2016, MNRAS, 458, 4250, doi: 10.1093/mnras/stw589

    Abarca, D. 2016, MNRAS, 458, 4250, doi: 10.1093/mnras/stw589

  42. [51]

    Steinberg, E., & Stone, N. C. 2024, Nature, 625, 463, doi: 10.1038/s41586-023-06875-y 20

  43. [52]

    1997, ApJ, 489, 87, doi: 10.1086/304783

    Storchi-Bergmann, T., Eracleous, M., Teresa Ruiz, M., et al. 1997, ApJ, 489, 87, doi: 10.1086/304783

  44. [53]

    2003a, in Astronomical Society of the Pacific Conference Series, Vol

    Storchi-Bergmann, T., Nemmen da Silva, R., & Eracleous, M. 2003a, in Astronomical Society of the Pacific Conference Series, Vol. 297, Star Formation Through Time, ed. E. Perez, R. M. Gonzalez Delgado, & G. Tenorio-Tagle, 431, doi: 10.48550/arXiv.astro-ph/0211477

  45. [54]

    S., Peterson, B

    Storchi-Bergmann, T., Schimoia, J. S., Peterson, B. M., et al. 2017, ApJ, 835, 236, doi: 10.3847/1538-4357/835/2/236

  46. [55]

    2003b, ApJ, 598, 956, doi: 10.1086/378938

    Storchi-Bergmann, T., Nemmen da Silva, R., Eracleous, M., et al. 2003b, ApJ, 598, 956, doi: 10.1086/378938

  47. [56]

    V., Strauss, M

    Strateva, I. V., Strauss, M. A., Hao, L., et al. 2003, AJ, 126, 1720, doi: 10.1086/378367

  48. [57]

    E., & Quataert, E

    Strubbe, L. E., & Quataert, E. 2009, MNRAS, 400, 2070, doi: 10.1111/j.1365-2966.2009.15599.x

  49. [58]

    E., & Quataert, E

    Strubbe, L. E., & Quataert, E. 2011, MNRAS, 415, 168, doi: 10.1111/j.1365-2966.2011.18686.x

  50. [59]

    2024, ApJ, 961, 172, doi: 10.3847/1538-4357/ad147d

    Ward, C., Gezari, S., Nugent, P., et al. 2024, ApJ, 961, 172, doi: 10.3847/1538-4357/ad147d

  51. [60]

    2022, A&A, 666, A6, doi: 10.1051/0004-6361/202142616

    Wevers, T., Nicholl, M., Guolo, M., et al. 2022, A&A, 666, A6, doi: 10.1051/0004-6361/202142616

  52. [61]

    Y., Webster, R

    Yong, S. Y., Webster, R. L., King, A. L., et al. 2017, PASA, 34, e042, doi: 10.1017/pasa.2017.37

  53. [62]

    2025, ApJ, 993, 244, doi: 10.3847/1538-4357/ae0581

    Zhang, M., Zhang, W., Deng, H., Guo, H., & Sun, J. 2025, ApJ, 993, 244, doi: 10.3847/1538-4357/ae0581

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.