Pith. sign in

REVIEW 5 major objections 4 minor 31 references

For most tidal disruption events, modeling the star's orbit as parabolic systematically underestimates the host black hole mass.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-02 17:24 UTC pith:HNKFIK2I

load-bearing objection A useful but incomplete extension of the Zhong (2025) model to 30 ZTF TDEs; the e-β diagnostic is new, but the mass-bias claim rests on an untested rigid-shift assumption and Eq. 4 has a typo. the 5 major comments →

arxiv 2603.25065 v2 pith:HNKFIK2I submitted 2026-03-26 astro-ph.HE

Fitting the light curves of tidal disruption events with non-parabolic model

Shiyan Zhong , Chenxi Shang , Xiaowei Liu This is my paper
classification astro-ph.HE
keywords tidal disruption eventsblack hole massesorbital eccentricitymass fallback ratelight curve fittingsupermassive black holesnuclear star clustersnon-parabolic model
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper claims that the standard assumption of a parabolic orbit in tidal disruption event (TDE) light-curve fitting is not just a numerical convenience but a source of systematic error. It constructs a model in which the disrupted star's specific orbital energy is a free parameter, shifting the debris energy distribution, and fits it to 30 optically discovered TDEs. The central result is a bias: parabolic fits underestimate black hole mass for eccentric orbits and overestimate it for hyperbolic orbits, with a derived scaling M_BH proportional to (1 - tilde_epsilon_orb)^(17/5). Most of the sample (27 out of 30) is inferred to be on eccentric orbits, implying that many black hole masses measured with parabolic fits may be too low. The paper also uses the eccentricity-penetration plane to argue that most events are consistent with two-body relaxation, with six outliers suggesting other dynamical channels.

Core claim

On the paper's own terms, the discovery is that the orbital energy of the disrupted star must be included in TDE light-curve modeling if black hole masses are to be trusted. The non-parabolic model replaces the parabolic debris mass distribution with a shifted version, adding the scaled orbital energy tilde_epsilon_orb as a fitting parameter. Applied to 30 events, the recovered black hole masses depend strongly on this parameter, following the analytic relation M_BH proportional to (1 - tilde_epsilon_orb)^(17/5); eccentric-orbit events need larger masses than parabolic fits give, while hyperbolic events need smaller ones. Since 90% of the sample has eccentricity less than unity, with deviati

What carries the argument

The central object is the shifted debris mass distribution, written as dm/depsilon(epsilon) = (dm/depsilon)_parabolic(epsilon - epsilon_orb). The paper parameterizes this using the scaled orbital energy tilde_epsilon_orb = epsilon_orb / Delta_epsilon, where Delta_epsilon is the typical half-width of the parabolic debris energy spread. This shift changes the fraction of bound debris and the timescale of the most-bound debris; by requiring the mass fallback rate to have the same peak and timescale under both models, the paper derives the scaling M_BH proportional to (1 - tilde_epsilon_orb)^(17/5).

Load-bearing premise

The entire analysis rests on the assumption that the mass distribution of debris from a non-parabolic disruption is exactly the parabolic distribution shifted rigidly by the star's orbital energy, with unchanged width and shape; if the energy spread or shape actually changes with orbital energy or penetration depth, the inferred orbital energies, eccentricities, and black hole mass biases would not be physical.

What would settle it

Run a hydrodynamical simulation of a star disrupted on an eccentric orbit with known parameters, apply the model's shifted-distribution equation to predict the debris spread, and check whether the recovered scaled orbital energy matches the input. Alternatively, take a TDE with an independently measured host black hole mass: if the event is classified as eccentric, the parabolic-fit mass should be lower than both the non-parabolic mass and the independent mass.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Parabolic-fit black hole masses for eccentric TDEs are biased low, with a correction factor on the order of (1 - tilde_epsilon_orb)^(-17/5).
  • If most TDEs are on slightly eccentric orbits, population-level black hole mass estimates from TDE light curves are likely shifted downward.
  • TDE light-curve fitting can now deliver orbital eccentricity measurements, opening a new observational window into loss-cone dynamics.
  • The eccentricity-penetration plane can separate standard two-body relaxation production from other mechanisms; the six high-penetration outliers motivate alternative formation channels.
  • For hyperbolic-orbit events, parabolic fits overestimate the black hole mass, so the sign of the bias depends on whether the star was bound or unbound.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Extending the paper's logic, previously published TDE-based black hole mass scaling relations may be systematically underestimated if TDEs are preferentially drawn from slightly bound orbits.
  • The same shifted-distribution machinery could unify single and repeated partial TDEs by treating orbital energy as a continuous parameter rather than fixing parabolicity.
  • A direct check would combine this model with independently measured host black hole masses for a few events: for an eccentric event, the parabolic mass should fall below the independent mass while the non-parabolic mass should agree.
  • If future hydrodynamic simulations show that the debris energy spread or shape depends on orbital energy or penetration depth, the recovered tilde_epsilon_orb values would become model artifacts rather than physical measurements.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 4 minor

Summary. The paper extends a previous fallback-based TDE light-curve model (Zhong 2025) by adding the scaled orbital energy of the disrupted star, tilde_epsilon_orb, as a free parameter. The non-parabolic debris distribution is constructed by rigidly shifting the parabolic dm/dε curve from the Guillochon & Ramirez-Ruiz (2013) library by ε_orb. The model is fit to 30 ZTF-I TDEs and compared with parabolic fits (tilde_epsilon_orb = 0). The authors claim that parabolic models systematically underestimate M_BH for eccentric events and overestimate it for hyperbolic events; that most of the sample (24/30 or ~90%) is on eccentric orbits with e < 1; and that six high-β, e<1 events cannot be explained by two-body relaxation in spherical nuclear star clusters. The paper also derives an analytic mass-biasing relation M_BH ∝ (1−tilde_epsilon_orb)^{17/5}.

Significance. If the rigid-shift assumption and the fits are valid, the paper would establish an important systematic correction for TDE black-hole mass measurements and would open the orbital eccentricity of disrupted stars as a new observable diagnostic for stellar dynamics in galactic nuclei. The use of a public fallback library, the explicit analytic scaling, and the W-AIC model comparison are useful and transparent elements. However, the central physical claim is conditional on one unvalidated ansatz, and there is a load-bearing algebraic inconsistency in the printed eccentricity mapping. The e−β population discussion is interesting but would need to be re-evaluated after the parameter mapping is fixed.

major comments (5)
  1. [§2, Eqs. (4) and (7); Table 2] The printed relation tilde_epsilon_orb = −β(1−e)/(2 q^{1/3}) is algebraically wrong. With r_t = r_*(M/m_*)^{1/3}, Δε = GM r_*/r_t^2, and ε_orb = −GM(1−e)β/(2 r_t), the correct expression is tilde_epsilon_orb = −β(1−e) q^{1/3}/2. Eq. (7) repeats the denominator error. Table 2 confirms the correct numerator form: AT 2018hco has log M=6.98, m_*=0.24, β=0.97, e=0.9983, which under the printed formula would give tilde ~ −2×10^{-6}, not the reported −0.28. The numerical example also states q=100 for a 10^6 M_sun BH and a 1 M_sun star; q should be 10^6. Since e and β are derived from this mapping, this must be corrected and the code/printed formulas reconciled before the fitted eccentricities can be trusted.
  2. [§2, Eq. (1)] The model consists of taking the parabolic debris energy distribution from Guillochon & Ramirez-Ruiz (2013) and rigidly shifting it by ε_orb, preserving the width and shape unchanged. No test from hydrodynamical or N-body simulations is given for how the debris energy distribution changes with the incoming orbital energy at fixed β, even though the same library shows strong β-dependent shape changes. If the width or asymmetry of dm/dε depends on ε_orb — for example because pericenter compression and shock heating depend on orbital energy — then the fitted tilde_epsilon_orb, e, β, and the claimed M_BH bias in §3.1 are properties of the ansatz, not of the observed TDEs. A validation run against existing simulation suites (e.g. Hayasaki et al. 2018) or an explicit controlled test should be added.
  3. [§3.1, Eq. (12) and Fig. 2] The relation M_BH ∝ (1−tilde_epsilon_orb)^{17/5} is derived from the same simplified uniform-dm/dε fallback model family that motivates the fits. The approximate agreement of the 30 data points with this curve in the lower panel of Fig. 2 is therefore partly a self-consistency check of the analytic degeneracy of the assumed fallback model, not independent empirical confirmation of the physical bias. The paper should label this as a model-consistency check and strengthen it with an external test, for example fitting synthetic light curves generated with a known input ε_orb and checking recovery.
  4. [§2, treatment of tilde_epsilon_orb] The text argues that ε_orb is dynamically fixed and independent of M_BH and m_*, yet the fitting parameter is tilde_epsilon_orb = ε_orb/Δε(M_BH, m_*) with a flat prior. This induces an M- and m-dependent effective prior on ε_orb itself. The inferred eccentricity distribution — including the 90%-e<1 claim in §3.1 — may therefore be prior-driven. Please justify this parameterization or report a robustness test where ε_orb is sampled directly or with a physically motivated prior.
  5. [§3, Tables 2–3 and W-AIC comparison] The model comparison is presented only through W-AIC scores. The reported W-AIC differences are small for many events, and the paper claims that the non-parabolic model is preferred only for a few objects. In this situation the mass-bias and eccentricity results could be sensitive to the intrinsic scatter term σ, the binning procedure, and the chosen prior ranges. It would strengthen the paper to show posterior distributions or at least a corner-plot example for representative events to demonstrate that the parameters are constrained by the data rather than by prior boundaries.
minor comments (4)
  1. [Table 2 note] The note misidentifies the columns: it says "the sixth column is the scaled penetration factor b and eighth column is the penetration factor. The eighth column is the W AIC score." The table layout shows b, β, and W-AIC in columns 6, 7, and 8; the note should be corrected.
  2. [§2.1, numerical example] For a 10^6 M_sun BH and a 1 M_sun star, q=10^6, not q=100. With the corrected q^{1/3} factor, the stated fractional difference of order 0.1 follows; as printed the numbers do not. Please fix the text and re-check other numerical statements that depend on this example.
  3. [Eq. (9)] Eq. (9) should use the corrected q^{1/3} factor from Eq. (4). If the printed Eq. (4) is corrected to q^{1/3} in the numerator, the right-hand side of Eq. (9) must be updated consistently.
  4. [References and figures] There are small formatting issues in the reference list (e.g., "V ." with an extra space in several author names). The figures are not embedded in the provided text, so their quality and axis labels could not be checked; please ensure they are complete in the submitted version.

Circularity Check

2 steps flagged

The central M_BH bias relation is an internal degeneracy of the rigid-shift model, and the e–β outlier classification leans on a load-bearing self-citation.

specific steps
  1. fitted input called prediction [Section 2.1 (Eqs. 10–13) and Section 3.1 (Fig. 2)]
    "The simple fallback rate (equation 6) is fully determined by t_min and mdot_peak. ... When fitting the same fallback rate – and consequently the same light curve under identical radiation processes – with both parabolic and non-parabolic models, the models must reproduce the same t_min and mdot_peak. Thus, the parabolic and non-parabolic models will yield different estimates for M_BH and m_*."

    The scaling relation M_BH ∝ (1−epsilon_tilde_orb)^{17/5} is derived by requiring t_min and mdot_peak to stay fixed when epsilon_tilde_orb changes in the same rigid-shift fallback model. Because epsilon_tilde_orb is a free parameter fitted to the same light curves as M_BH, the correlation displayed in Fig. 2 is built into the likelihood by construction. The quoted 'must reproduce the same t_min and mdot_peak' condition is exactly the degeneracy that forces the fitted M_BH to change with epsilon_tilde_orb. Thus the Section 3.1 agreement with Eq. (12) is a self-consistency check of the model's own degeneracy, not an independent empirical confirmation of the claim that parabolic fits underestimate M_BH for eccentric events. The sign and amplitude of the claimed bias are already encoded in Eq.

  2. self citation load bearing [Section 3.2, Eq. (14) and Fig. 4 discussion]
    "From equation 23 of Zhong et al. (2023), the maximum eccentricity is ... Zhong et al. (2023) demonstrated that in a realistic (but spherical) NSC, nearly all full TDEs in the e < 1 quadrant should have beta = 1 ... The other 6 TDEs lie beyond the boundary beta = beta_d (for e < 1) derived in Zhong et al. (2023)."

    The central observational claim that six TDEs are outliers requiring 'alternative mechanisms' depends on a theoretical boundary (e_max and beta ≈ beta_d) imported from the authors' own prior paper. The e and beta values being classified were themselves produced by the non-parabolic model built on the authors' earlier Zhong (2025) framework. The boundary is load-bearing: without adopting the self-cited Zhong et al. (2023) curve, the 6-outlier classification does not follow. This is not a uniqueness theorem, but it is a self-citation used as the decisive criterion for the paper's dynamical-origin conclusions.

full rationale

The paper's core derivation is not externally circular in the sense of fitting to a benchmark and then predicting that benchmark: the light curves come from the independent ZTF-I sample. However, the central physical conclusion—that non-parabolic fits imply a systematic M_BH correction scaling as (1−epsilon_tilde_orb)^{17/5}—is a direct consequence of the same rigid-shift ansatz (Eq. 1) used both to build the model and to measure epsilon_tilde_orb. The Section 3.1 agreement with Eq. (12) is therefore a demonstration that the fitting machinery reproduces the model's own analytic degeneracy, not a test against independent eccentricity measurements or hydrodynamic simulations. The Section 3.2 interpretation of the e–β plane also rests on the authors' previous Zhong et al. (2023) boundary, which is self-cited and load-bearing for the six-outlier claim. These features make the paper partially circular (score 6), even though the model is stated transparently and the authors acknowledge deviations from the simple relation. No evidence of renaming known results or importing a uniqueness theorem was found beyond the self-cited boundary.

Axiom & Free-Parameter Ledger

6 free parameters · 4 axioms · 0 invented entities

The central claim rests on one new fitted parameter (tilde_epsilon_orb), the standard MOSFiT fitting parameters, and several domain assumptions inherited from prior work. No new physical entities are postulated. The main unvalidated assumption is the rigid shift of the debris energy distribution.

free parameters (6)
  • tilde_epsilon_orb (scaled orbital energy) = -0.93 to +0.08 across sample (Table 2)
    The new parameter defining the non-parabolic model; its fitted value drives the claimed M_BH shift and the derived eccentricity.
  • M_BH (black hole mass) = log10(M_BH/M_sun) ~ 6.2-7.6 (Table 2/3)
    The key inferred quantity; the paper claims parabolic fits bias it.
  • m_star (stellar mass) = 0.1-11 M_sun (Table 2)
    Fitted with Kroupa prior; enters Δε and the e-β relation.
  • b (scaled penetration factor) = 0.38-1.62 (Table 2)
    Fitted; required to convert tilde_epsilon_orb to eccentricity and β.
  • Radiation parameters (η, R_ph0, l, t_nu) = not listed individually
    Standard MOSFiT-like photosphere parameters needed to map fallback rate to light curves; affect the fit but not the central bias claim directly.
  • n_H and σ = not listed individually
    Extinction and error-scaling parameters; standard fitting nuisance parameters.
axioms (4)
  • domain assumption The non-parabolic debris mass distribution is the parabolic dm/dε shifted rigidly by ε_orb (Eq. 1), with unchanged width and shape.
    Core model assumption inherited from Zhong (2025); not tested against hydrodynamic simulations for large |ε_orb|.
  • domain assumption Kepler's third-law fallback conversion dε/dt = (1/3)(2πGM)^{2/3} t^{-5/3} applies even for slightly hyperbolic stars.
    For ε_orb > 0 some debris may be unbound and never return; the formula assumes bound Keplerian return.
  • domain assumption The MOSFiT luminosity-dependent photosphere model correctly converts fallback rate to multi-band light curves.
    Adopted without re-validation; the radiation model is known to be a simplified approximation.
  • domain assumption The e-β boundary for TDE origin from Zhong et al. (2023) is correct and applicable to this sample.
    Used to classify 24/30 events as consistent with two-body relaxation and to flag six outliers; this is a self-cited theoretical framework.

pith-pipeline@v1.3.0-alltime-deepseek · 14010 in / 12818 out tokens · 126917 ms · 2026-08-02T17:24:41.374849+00:00 · methodology

0 comments
read the original abstract

Tidal disruption events (TDEs) are powerful probes of supermassive black hole (SMBH) properties and accretion physics. The existing light curve fitting tools assume that the disrupted stars are on parabolic orbits, which may introduce systematic biases in derived parameters. In this work, we extend the model of Zhong (2025) to construct a non-parabolic TDE model that incorporates orbital energy of the disrupted star as a free parameter ($\tilde{\epsilon}_{\rm orb}$) to modify the debris mass distribution and mass fallback rate. We apply this model to 30 TDEs from the ZTF-I survey and compare the results with those from a standard parabolic model. We find that neglecting orbital energy leads to biased black hole mass estimates: for eccentric (hyperbolic) orbits, parabolic models systematically underestimate (overestimate) the black hole mass. Additionally, we measure orbital eccentricities ($e$) and penetration factors ($\beta$) of the disrupted stars in this sample, enabling an investigation of their origins via the $e$-$\beta$ parameter space. Most events (24/30) are consistent with production via two-body relaxation in spherical nuclear star clusters, but six outliers with high $\beta$ and $e<1$ suggest alternative mechanisms. Our results highlight the importance of accounting for orbital energy in TDE modeling to improve the accuracy of SMBH mass measurements and to better understand the dynamical origin of the disrupted stars.

Figures

Figures reproduced from arXiv: 2603.25065 by Chenxi Shang, Shiyan Zhong, Xiaowei Liu.

Figure 1
Figure 1. Figure 1: Comparison of BH masses obtained from parabolic TDE model ( [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The dependence of the fitted BH mass difference on the orbital eccentricity (upper panel) and [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The locations of the 30 TDEs on the e-β plane. The theoretical model of Zhong et al. (2023) assumes a spherically symmetric nuclear star cluster (NSC). However, real NSCs may exhibit deviations from spherical symmetry; for instance, bulk rotation has been observed in the Milky Way’s NSC, leading to a flattened, axisymmetric structure (Feldmeier et al. 2014). In an axisymmetric potential, angular momentum i… view at source ↗
Figure 4
Figure 4. Figure 4: The positions of the 8 full TDEs with β > βd in the e-β plane. The vertical black lines indicate the range of βd, estimated from the posterior mass distribution of the disrupted star (see, equations 1–5 of Zhong (2025)). The eccentricities of AT 2018lni and AT 2019teq are consistent with e = 1 (albeit with large uncertainties), so they are permitted by the model of Zhong et al. (2023). The other 6 TDEs lie… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

31 extracted references · 1 linked inside Pith

  1. [1]

    2024, ApJ, 977, 279 2

    Bao, D.-W., Guo, W.-J., Zhang, Z.-X., et al. 2024, ApJ, 977, 279 2

    2024

  2. [2]

    Cohn, H., & Kulsrud, R. M. 1978, ApJ, 226, 1087 8, 11 14 Zhong, Shang, Liu

    1978

  3. [3]

    2014, A&A, 570, A2 12

    Feldmeier, A., Neumayer, N., Seth, A., et al. 2014, A&A, 570, A2 12

    2014

  4. [4]

    W., Lang, D., & Goodman, J

    Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP, 125, 306 4

    2013

  5. [5]

    Frank, J., & Rees, M. J. 1976, MNRAS, 176, 633 2, 8

    1976

  6. [6]

    2020, MNRAS, 497, 1925 2

    Gomez, S., Nicholl, M., Short, P., et al. 2020, MNRAS, 497, 1925 2

    2020

  7. [7]

    A., et al

    Guillochon, J., Nicholl, M., Villar, V . A., et al. 2018, ApJS, 236, 6 2

    2018

  8. [8]

    2013, ApJ, 767, 25 3, 8, 9, 10

    Guillochon, J., & Ramirez-Ruiz, E. 2013, ApJ, 767, 25 3, 8, 9, 10

    2013

  9. [9]

    2023, ApJ, 942, 9 2, 6, 12

    Hammerstein, E., van Velzen, S., Gezari, S., et al. 2023, ApJ, 942, 9 2, 6, 12

    2023

  10. [10]

    2018, ApJ, 855, 129 2, 8

    Hayasaki, K., Zhong, S., Li, S., Berczik, P., & Spurzem, R. 2018, ApJ, 855, 129 2, 8

    2018

  11. [11]

    T., Auchettl, K., Hoogendam, W

    Hinkle, J. T., Auchettl, K., Hoogendam, W. B., et al. 2024, arXiv e-prints, arXiv:2412.15326 2 Kov´acs-Stermeczky, Z. V ., & Vink´o, J. 2023a, PASP, 135, 034102 2 Kov´acs-Stermeczky, Z. V ., & Vink´o, J. 2023b, PASP, 135, 104102 2

    Pith/arXiv arXiv 2024

  12. [12]

    Law-Smith, J. A. P., Coulter, D. A., Guillochon, J., Mockler, B., & Ramirez-Ruiz, E. 2020, ApJ, 905, 141 9

    2020

  13. [13]

    P., & Shapiro, S

    Lightman, A. P., & Shapiro, S. L. 1977, ApJ, 211, 244 8

    1977

  14. [14]

    2024, ApJ, 971, L26 2

    Lin, Z., Jiang, N., Wang, T., et al. 2024, ApJ, 971, L26 2

    2024

  15. [15]

    2025, ApJ, 987, L20 2

    Makrygianni, L., Arcavi, I., Newsome, M., et al. 2025, ApJ, 987, L20 2

    2025

  16. [16]

    2019, ApJ, 872, 151 2, 4

    Mockler, B., Guillochon, J., & Ramirez-Ruiz, E. 2019, ApJ, 872, 151 2, 4

    2019

  17. [17]

    2021, ApJ, 906, 101 2

    Mockler, B., & Ramirez-Ruiz, E. 2021, ApJ, 906, 101 2

    2021

  18. [18]

    R., et al

    Nicholl, M., Wevers, T., Oates, S. R., et al. 2020, MNRAS, 499, 482 2

    2020

  19. [19]

    Rees, M. J. 1988, Nature, 333, 523 1

    1988

  20. [20]

    Ryu, T., Krolik, J., Piran, T., & Noble, S. C. 2020, ApJ, 904, 100 9

    2020

  21. [21]

    J., Ravi, V ., Yao, Y ., et al

    Somalwar, J. J., Ravi, V ., Yao, Y ., et al. 2025, ApJ, 985, 175 2

    2025

  22. [22]

    2025, ApJ, 982, 150 2

    Sun, J., Guo, H., Gu, M., et al. 2025, ApJ, 982, 150 2

    2025

  23. [23]

    2024, A&A, 692, A262 2

    Sun, L., Jiang, N., Dou, L., et al. 2024, A&A, 692, A262 2

    2024

  24. [24]

    C., & Ostriker, J

    Teboul, O., Stone, N. C., & Ostriker, J. P. 2024, MNRAS, 527, 3094 11

    2024

  25. [25]

    2013, ApJ, 774, 87 12

    Vasiliev, E., & Merritt, D. 2013, ApJ, 774, 87 12

    2013

  26. [26]

    2025, arXiv e-prints, arXiv:2510.26561 2

    Wang, Y ., Wang, T., Huang, S., et al. 2025, arXiv e-prints, arXiv:2510.26561 2

    arXiv 2025

  27. [27]

    2017, MNRAS, 468, 1760 11

    Weissbein, A., & Sari, R. 2017, MNRAS, 468, 1760 11

    2017

  28. [28]

    R., Pasham, D

    Wevers, T., Coughlin, E. R., Pasham, D. R., et al. 2023, ApJ, 942, L33 2

    2023

  29. [29]

    2025, ApJ, 983, 131 2, 3, 4, 13

    Zhong, S. 2025, ApJ, 983, 131 2, 3, 4, 13

    2025

  30. [30]

    2023, ApJ, 959, 19 3, 8, 9, 10, 12, 13

    Zhong, S., Hayasaki, K., Li, S., Berczik, P., & Spurzem, R. 2023, ApJ, 959, 19 3, 8, 9, 10, 12, 13

    2023

  31. [31]

    2025, Research in Astronomy and Astrophysics, 25, 015017 2

    Zhong, S., Xu, X., Chen, X., et al. 2025, Research in Astronomy and Astrophysics, 25, 015017 2

    2025