Pith. sign in

REVIEW 4 major objections 4 minor 60 references

The Rotating Bulge and Halo in the Milky Way: Evidence of Angular Momentum Transferred from the Decelerating Bar

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read The paper claims that the Milky Way's decelerating bar, not accretion or initial conditions, spun up the stars that form the observed rotating bulge and inner halo.

desk verdict A plausible, readable case that a decelerating bar spins up the bulge and inner halo, but the control run does not cleanly isolate deceleration from bar growth, so the causal claim is provisional. read the letter →

arxiv 2506.12717 v1 pith:CRFUJUJ3 submitted 2025-06-15 astro-ph.GA

classification astro-ph.GA
keywords GalacticbarbulgeMilkyWaystellarhaloangularmomentumtransferdecelerationtest-particlesimulationneuralnetworkclassificationGaiaDR3
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 identifies a population of 1,175,737 stars in Gaia DR3 whose neural-network scores mark them as neither ordinary disk nor obvious accreted debris, and shows they rotate at roughly $v_\phi\simeq80\,\mathrm{km\,s^{-1}}$. It then argues that this rotation is not primordial: in a test-particle simulation, a central bar whose pattern speed $\Omega_b$ falls from $-56$ to $-35\,\mathrm{km\,s^{-1}}\,\mathrm{kpc}^{-1}$ over 4 Gyr transfers angular momentum to bulge and halo stars, reproducing the observed $v_\phi(R)$ profile, while a steadily rotating bar does not. If correct, the Milky Way's bulge and inner halo acquired their net prograde motion secularly, through dynamical friction braking the bar, rather than through initial conditions or accretion alone. The result would make the decelerating bar a central engine for the inner Galaxy's kinematics.

What carries the argument

The central object is the decelerating bar potential: a simple analytic bar whose pattern speed $\Omega_b$ drops from $-56$ to $-35\,\mathrm{km\,s^{-1}}\,\mathrm{kpc}^{-1}$ (about 37.5% deceleration) while its mass grows by a factor of 2.0 and its radial extent by a factor of 1.26 over 4 Gyr. The test particles are pseudo-stars sampled from an equilibrium distribution-function model of the Milky Way, evolved in the axisymmetric background plus the bar. The transfer is tracked through changes in angular momentum as a function of the resonance ratio $\epsilon=(\Omega_\phi-\Omega_b)/\Omega_r$; particles trapped near corotation ($\epsilon\simeq0$) gain the most angular momentum, and the neural-network-selected bulge and halo subset shows net gains that a steadily rotating bar cannot reproduce.

What would settle it

Run the control simulation with a bar whose pattern speed is constant at $-35\,\mathrm{km\,s^{-1}}\,\mathrm{kpc}^{-1}$ but whose mass and radial extent grow exactly as in the decelerating run (mass $\times2.0$, radius $\times1.26$ over 4 Gyr). If that control reproduces the observed $v_\phi(R)$ profile, the claim that deceleration is the cause fails; if it does not, deceleration is supported. An independent check would measure the present-day bar pattern speed from stellar kinematics and verify that it has fallen by roughly 37.5% over the last 4 Gyr, matching the assumed braking history.

Watch

Extended reading notes

Core claim

The paper's central claim is that the observed rotating component is not a distinct stellar population but a mixture of bulge, halo, and thick-disk stars that have been given angular momentum by the Milky Way's decelerating bar. The evidence is a test-particle simulation that initializes pseudo-stars from an equilibrium distribution-function model of the Galaxy and evolves them for 4 Gyr in an axisymmetric background plus a bar whose pattern speed drops by about 37.5% while its mass and radial extent grow. After applying observational errors, selection effects, and the same neural network used on the data, the simulated $v_\phi(R)$ profile for the selected stars agrees with the observed profile, and the bulge and halo components individually show the rotation seen in the data. A comparison run with a steadily rotating bar produces a profile that deviates significantly from observation. The paper concludes that dynamical friction decelerating the bar is the pivotal process shaping the inner Galaxy's kinematics.

Load-bearing premise

The whole case for deceleration rests on the steadily rotating bar comparison run: if that control did not include the same factor-2.0 mass growth and factor-1.26 radial growth as the main run, then the difference between the two curves could be caused by bar growth rather than by the change in pattern speed, and the paper does not state that the control includes that growth.

Editorial extensions

If this is right

  • If the decelerating bar is the source, the inner Galaxy's prograde rotation is still being built today, and the bulge and inner halo should gain angular momentum as long as the bar keeps braking.
  • The neural-network-selected stars should be predominantly bulge, halo, and thick-disk stars, a prediction that can be checked directly with elemental abundances and stellar ages.
  • The characteristic $v_\phi(R)$ shape, rising to about 3 kpc, falling to the solar circle, and rising again in the outer halo, is a fingerprint that future astrometric and spectroscopic surveys can look for.
  • Bar deceleration of roughly 37.5% over 4 Gyr places the Milky Way's bar in the slow-bar regime, so resonance trapping should be visible as clustered angular-momentum gains for halo stars beyond the solar radius.

Reading between the lines

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

  • If deceleration is the active ingredient, a control run in which the bar grows in mass and length while keeping a constant pattern speed should still fail to match the data; the paper does not state that its steady control includes that growth, so this is the natural next test.
  • Because the simulation neglects self-gravity, the real bar may transfer angular momentum even more efficiently than modeled, making the simulated rotation a possible lower bound on the bulge and halo spin.
  • The same torquing mechanism could explain why even metal-poor inner-halo stars rotate: the bar acts on pre-existing old stars, so no separate accretion origin is required.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper identifies a subset of 1,175,737 stars in Gaia DR3 with neural network predictions between 0.24 and 0.4 as an 'atypical rotating component' with net rotation of roughly 80 km/s. To explain its origin, the authors build a test-particle simulation with an axisymmetric distribution-function model of the Milky Way (following Binney & Vasiliev 2023, 2024) plus a central bar that decelerates from 56 to 35 km/s/kpc over 4 Gyr while also growing in mass and radial extent. They resample the simulation to mimic Gaia selection effects, apply the same neural network, and compare the simulated v_phi(R) profile of the selected particles with the observed one, reporting strong agreement. A comparison run with a steadily rotating bar shows a discrepancy, which the authors interpret as evidence that the decelerating bar transfers angular momentum to bulge and halo stars. The paper concludes that the rotating component is predominantly bulge, halo, and thick disk stars and that bar deceleration is pivotal in shaping the inner Galaxy's kinematics.

Significance. If the causal attribution holds, the paper would supply a large-sample observational check on a long-standing theoretical expectation: that a decelerating bar can secularly pump angular momentum into the bulge and inner stellar halo. The computational setup is well chosen in several respects: the simulation uses a physically motivated distribution-function model, includes synthetic Gaia-like observational errors and selection functions, and makes an explicit comparison to a steady-bar control. The authors also candidly list caveats about neglecting self-gravity and spiral arms. However, the central claim currently rests on a single prescribed bar model and a hand-selected neural network window, and the comparison in Figure 4 is visual rather than statistical. The paper is a useful contribution, but the deceleration hypothesis is not yet demonstrated at the strength claimed in the title and conclusions.

major comments (4)
  1. [Section 4.2, steady-bar control] The comparison that isolates bar deceleration is not fully specified. The main run has a bar that simultaneously decelerates (Omega_b from -56 to -35 km/s/kpc) and grows in mass and radial extent (Section 3.1: reaching 2.0 and 1.26 times initial values), but the control run is described only as maintaining a constant pattern speed Omega_b = -35 km/s/kpc. If the control omits the mass and radial growth, the difference between the blue dashed and solid curves in Figure 4 could be caused by bar growth rather than by deceleration. Since the title and Section 5 attribute the net rotation to the decelerating bar, the manuscript must specify the control run's mass and size evolution or rerun it with identical growth, and show that the discrepancy persists.
  2. [Section 4.2, Figure 4] The agreement between simulated and observed v_phi(R) profiles is asserted visually, with no quantitative goodness-of-fit, uncertainty estimate, or test of the difference between the decelerating and steady runs. Figure 4 shows median and 16th/84th percentile bins, but the model has many free parameters (DF component masses and action scales, bar parameters, selection thresholds), and no posterior or parameter variation is presented. A bootstrap or chi-squared comparison, and ideally a small exploration of the neural network prediction window, would be needed to support the claim of 'strong agreement' and the attribution of the rotation to bar deceleration.
  3. [Section 2.2, NN selection window] The definition of the rotating component depends on the hand-selected NN prediction interval [0.24, 0.4], with the right boundary chosen to exclude GSE stars. The subsequent composition analysis, Figure 3, and Figure 4 all use this specific window. The manuscript does not test the sensitivity of the v_phi(R) profile or the simulated-versus-observed agreement to the boundaries. If the profile is robust to these choices, that should be demonstrated; if not, the selection is a potential source of the claimed signal.
  4. [Section 3 and Appendix A] The initial distribution function is constrained by the same APOGEE/Gaia-based data that define the observed sample (the BV24 fits listed in Tables B1 and B2), so the simulation is not fully independent of the observations. The authors should discuss the extent to which the initial conditions already encode rotating bulge or halo kinematics, and clarify whether the bulge and halo DFs permit net rotation at t=0. The internal decelerating-versus-steady comparison is the strongest guard against circularity, but the absolute agreement in Figure 4 is not an independent confirmation.
minor comments (4)
  1. [Section 3.1, sign convention] The pattern speeds are given as negative numbers (Omega_b = -56 and -35 km/s/kpc), but the direction convention is not defined; the resonance variable epsilon = (Omega_phi - Omega_b)/Omega_r in Figure 5 assumes a sign convention that should be stated explicitly.
  2. [Section 5, 'for the first time'] The sentence 'For the first time, we confirmed the net rotation of both the bulge and inner halo on a million-level sample' is stronger than the cited literature warrants, given earlier reports of bulge and inner-halo rotation cited in the introduction; suggest softening to 'on a million-level sample' without the 'first time' phrasing.
  3. [Section 2.2, GSE exclusion] The right boundary of the selected NN interval is said to exclude GSE member stars, but no quantitative criterion is given; providing the actual boundary test or a reference would reduce the impression of arbitrariness in the sample definition.
  4. [Figure 4 caption] The caption text 'The solid lines illustrate a comparison of all simulated samples with NN prediction values ranging between 0.24 and 0.4' is a little unclear; the legend label 'SimSteady' is not described in the caption, and the labels for the orange and red dashed curves would be easier to follow if the figure legend and caption were aligned.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the v-phi profile is an emergent simulation output, not a fitted target; self-citations are methodological inputs, and the steady-bar control confound is a validity concern, not circularity.

full rationale

The paper is a forward modeling study rather than a fit-to-target derivation. The bar parameters (pattern-speed history from -56 to -35 km/s/kpc, mass growth x2.0, radial growth x1.26) are prescribed inputs taken from prior observational constraints and previous bar models, not adjusted to reproduce the observed v_phi(R) of the NN-selected component. The initial pseudo-star distribution comes from the pre-existing BV23/BV24 action-space DF calibrated to APOGEE data; it is not fitted to the specific 1,175,737-star sample or to its rotation curve. The neural network classifier is the authors' prior model (Li24), but it is applied identically to observations and simulations and its output does not directly encode v_phi; the 0.24-0.4 selection window is data-driven but does not by itself force the simulated rotation profile. The key causal comparison between the decelerating-bar run and the 'steadily rotating bar' run is an internal control, and the angular-momentum gains in Figure 5 are emergent, not imposed. Self-citations to Li24 and Li et al. (2023, 2024a) are methodological inputs and are not used to assert uniqueness or to forbid alternative explanations. The reader's/skeptic's concern that the steady-bar control may not include the same bar mass and radial growth as the main run is a legitimate potential confound in the causal attribution to deceleration, but it is a correctness/validity issue rather than circularity: nothing in the derivation is equivalent to its inputs by construction. Hence the circularity score is low.

Assumptions & free parameters 9 free parameters · 5 assumptions · 0 invented entities

The central claim rests on a chain of model assumptions: the initial DF, the neural network transferability, the prescribed bar evolution, and the selection function model. The free parameters are mostly carried over from prior fits rather than fitted here, but the hand-selected NN thresholds and the possibly unmatched control run are load-bearing choices that are not tested for sensitivity.

free parameters (9)
  • NN selection lower threshold = 0.24
    Chosen as the valley bottom left of the secondary peak in the prediction distribution (Section 2.2, Figure 1). The sample and all conclusions depend on this boundary.
  • NN selection upper threshold = 0.4
    Chosen to exclude GSE stars, hand-selected from the observed prediction distribution (Section 2.2).
  • Bar initial pattern speed = -56 km/s/kpc
    Prescribed from literature estimates (Chiba and Schönrich 2021; Li et al. 2023), not fitted to the target v_phi profile.
  • Bar final pattern speed = -35 km/s/kpc
    Prescribed to match present-day slow-bar estimates; used in both the decelerating and steady models.
  • Bar mass growth factor = 2.0
    Prescribed to mimic bar growth over 4 Gyr; possibly differing in the steady control, which would confound deceleration with growth.
  • Bar radial growth factor = 1.26
    Prescribed alongside mass growth; same confounding concern as the mass growth factor.
  • Bar phase angle = 28 degrees
    Chosen from present-day bar morphology observations (Wegg et al. 2015).
  • Simulation duration = 4 Gyr
    Chosen; angular momentum gains are endpoint-dependent and no convergence test is shown.
  • DF component masses and action parameters = Tables B1 and B2
    Taken from Binney and Vasiliev (2023, 2024) fits to APOGEE and other data; the initial pseudo-star population inherits these fitted values.
assumptions (5)
  • domain assumption The equilibrium axisymmetric Milky Way distribution function model (BV23/BV24) adequately represents the initial Galaxy.
    The initial pseudo-stars are sampled from this DF; if the DF is wrong, the bar's effect on the selected sample is not representative. Used throughout Section 3 and Appendix A.
  • domain assumption The neural network trained by Li24 maps kinematic inputs to a probability whose values correlate with stellar component in a way that transfers from FIRE-2 to the Milky Way and to this simulation.
    Used in Sections 2.1 and 4.1; both the observational sample selection and the simulation comparison rely on the NN score being meaningful for component separation.
  • domain assumption A rigid, prescribed decelerating bar in a fixed background potential is a sufficient approximation for angular momentum transfer to baryons over 4 Gyr.
    The bar's slowdown and growth are imposed rather than produced self-consistently by dynamical friction; self-gravity of the disk is neglected (Section 3.1).
  • domain assumption Observational uncertainties and the Gaia selection function are adequately captured by Gaussian and log-normal errors plus 3D resampling.
    Section 3.2; if the selection model is wrong, the simulated sample composition and v_phi profile are not directly comparable to the observed NN-selected sample.
  • standard math Standard stellar-dynamical background results (action-angle coordinates, Jeans theorem, AGAMA orbit integration) are taken as given.
    Used throughout for distribution function modeling and orbital integration.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Rotating Bulge and Halo in the Milky Way: Evidence of Angular Momentum Transferred from the Decelerating Bar." pith.science (2026). https://pith.science/paper/CRFUJUJ3

@misc{pith2026250612717,
  author       = {Pith},
  title        = {Pith review of: The Rotating Bulge and Halo in the Milky Way: Evidence of Angular Momentum Transferred from the Decelerating Bar},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CRFUJUJ3}},
  note         = {Machine review of arXiv:2506.12717}
}
read the original abstract

Recent observations indicate that both the Milky Way bulge and inner halo exhibit angular momentum, although the origin and evolution of this prograde signature remain ambiguous. One plausible scenario involves secular evolution induced by the central bar and spiral arms. In this study, we identified a component consisting of 1,175,737 stars with net rotation through the application of a neural network (NN) method. To investigate the composition of this rotating sample and the origin of its rotation, we conducted a test particle simulation incorporating an equilibrium axisymmetric background potential together with a central decelerating bar. The test particles were generated using a distribution function (DF) model derived from observational constraints. Our results indicate that the decelerating bar transfers angular momentum to the pseudo-stars, and the rotational profile from our simulation shows strong agreement with observational data. These findings suggest that the rotating sample identified by our NN model predominantly comprises bulge, halo, and thick disk stars, and that the central decelerating bar is pivotal in shaping the inner Galaxy's kinematics through angular momentum transfer.

Figures

Figures reproduced from arXiv: 2506.12717 by the authors.

Figure 1
Figure 1. Upper panel: neural network predictions for the Gaia DR3 sample. We show the kernel density distribution on both linear and logarithmic scales. The dashed line marks the threshold in Li et al. (2024b). The area of interest for this study, specifically the range with prediction values be￾tween 0.24 and 0.4, is highlighted with a shaded background. Lower panel: neural network predictions for the resampled simulation d… view at source ↗
Figure 2
Figure 2. Upper panel: the m=0 and m=2 Fourier terms as a function of distance at z = 0 for the slowing down bar model in black and red respectively. Lower panel: the pat￾tern speed Ωb in this work is shown in solid black curve, and the decelerating rate Ω˙ b in this work is shown in dashed black curve. In order to make a comparison, the pattern speed Ωb and the decelerating rate Ω˙ b in Li et al. (2024a) are shown in red sol… view at source ↗
Figure 3
Figure 3. Distribution in the coordinate space for stars with NN predictions in the selected range. The upper row represents observational data, while the lower row depicts simulation results. Each subplot is color-coded to indicate the number density of stars, with warmer colors signifying higher densities. We analyze the variation of vϕ with R for both sim￾ulated and observed samples, as illustrated in [PITH_FULL_IMAGE:fig… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: The variation of vϕ with R for simulated and observed samples. The dots represent the median Vϕ val￾ues in each bin, with error bars indicating the 16th and 84th percentiles. The solid lines illustrate a comparison of all sim￾ulated samples with NN prediction values ra…
Figure 5
Figure 5. Figure 5: Depiction of the angular momentum transfer for bulge stars and halo stars with prediction values between 0.24 and 0.4. The left column is color-coded by the number density, while the right column is color-coded to indicate the Galactocentric distance. The parameters de…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 14 canonical work pages

  1. [1]

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

    ENTRY address archivePrefix author booktitle chapter doi edition editor eprint howpublished institution journal key month number organization pages publisher school series title misctitle type volume year version url label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts ...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION format.url url empty "" new.block "" url * "" * if FUNCTION format.eprint eprint empty "" archivePrefix empty "" archivePrefix "arXiv" = new.block " " eprint * " " * new.block " " eprint * " " * if if if FUNCTION format.doi doi empty "" " " doi * " " * if FUNCTION format.pid doi empty eprint empty ur...

  3. [3]

    PH, 1 h 8SYġ0!@Xqboߣp=@:ߡcce& k<q b; ;< u s!p9w

    thebibliography [1] 20pt to REFERENCES 6pt =0pt -12pt 10pt plus 3pt =0pt =0pt =1pt plus 1pt =0pt =0pt -12pt =13pt plus 1pt =20pt =13pt plus 1pt \@M =10000 =-1.0em =0pt =0pt 0pt =0pt =1.0em @enumiv\@empty 10000 10000 `\.\@m \@noitemerr \@latex@warning Empty `thebibliography' environment \@ifnextchar \@reference \@latexerr Missing key on reference command E...

  4. [4]

    Ablimit , I., Zhao , G., Flynn , C., & Bird , S. A. 2020, , 895, L12, 10.3847/2041-8213/ab8d45

  5. [5]

    Anders, F., Khalatyan, A., Queiroz, A. B. d. A., et al. 2022, Astronomy & Astrophysics, 658, A91

  6. [6]

    Ardern-Arentsen , A., Monari , G., Queiroz , A. B. A., et al. 2024, , 530, 3391, 10.1093/mnras/stae1049

  7. [7]

    W., Koposov , S

    Belokurov , V., Erkal , D., Evans , N. W., Koposov , S. E., & Deason , A. J. 2018, , 478, 611, 10.1093/mnras/sty982

  8. [8]

    2019, , 482, 1417, 10.1093/mnras/sty2813

    Bennett , M., & Bovy , J. 2019, , 482, 1417, 10.1093/mnras/sty2813

Show all 60 references
  1. [9]

    2014, , 440, 787, 10.1093/mnras/stu297

    Binney , J. 2014, , 440, 787, 10.1093/mnras/stu297

  2. [10]

    2008, Galactic Dynamics: Second Edition

    Binney , J., & Tremaine , S. 2008, Galactic Dynamics: Second Edition

  3. [11]

    2023, , 520, 1832, 10.1093/mnras/stad094

    Binney , J., & Vasiliev , E. 2023, , 520, 1832, 10.1093/mnras/stad094

  4. [12]

    2024, , 527, 1915, 10.1093/mnras/stad3312

    ---. 2024, , 527, 1915, 10.1093/mnras/stad3312

  5. [13]

    1943, , 97, 255, 10.1086/144517

    Chandrasekhar , S. 1943, , 97, 255, 10.1086/144517

  6. [14]

    2023, , 525, 3576, 10.1093/mnras/stad2324

    Chiba , R. 2023, , 525, 3576, 10.1093/mnras/stad2324

  7. [15]

    Chiba , R., Friske , J. K. S., & Sch \"o nrich , R. 2021, , 500, 4710, 10.1093/mnras/staa3585

  8. [16]

    2021, , 505, 2412, 10.1093/mnras/stab1094

    Chiba , R., & Sch \"o nrich , R. 2021, , 505, 2412, 10.1093/mnras/stab1094

  9. [17]

    2022, , 513, 768, 10.1093/mnras/stac697

    ---. 2022, , 513, 768, 10.1093/mnras/stac697

  10. [18]

    P., & Gerhard , O

    Clarke , J. P., & Gerhard , O. 2022, , 512, 2171, 10.1093/mnras/stac603

  11. [19]

    P., & Sellwood , J

    Debattista , V. P., & Sellwood , J. A. 2000, , 543, 704, 10.1086/317148

  12. [20]

    M., Belokurov , V., Evans , N

    Dillamore , A. M., Belokurov , V., Evans , N. W., & Davies , E. Y. 2023, , 524, 3596, 10.1093/mnras/stad2136

  13. [21]

    2022, arXiv e-prints, arXiv:2205.15725, 10.48550/arXiv.2205.15725

    Dootson , D., & Magorrian , J. 2022, arXiv e-prints, arXiv:2205.15725, 10.48550/arXiv.2205.15725

  14. [22]

    W., Rix , H.-W., & Ness , M

    Eilers , A.-C., Hogg , D. W., Rix , H.-W., & Ness , M. K. 2019, , 871, 120, 10.3847/1538-4357/aaf648

  15. [23]

    2016, , 595, A1, 10.1051/0004-6361/201629272

    Gaia Collaboration , et al. 2016, , 595, A1, 10.1051/0004-6361/201629272

  16. [24]

    2023, , 674, A1, 10.1051/0004-6361/202243940

    ---. 2023, , 674, A1, 10.1051/0004-6361/202243940

  17. [25]

    A., Arzamasskiy , L., & Duarte , V

    Hamilton , C., Tolman , E. A., Arzamasskiy , L., & Duarte , V. N. 2023, , 954, 12, 10.3847/1538-4357/acd69b

  18. [26]

    2018, The Astrophysical Journal, 863, 113

    Haywood, M., Di Matteo, P., Lehnert, M., et al. 2018, The Astrophysical Journal, 863, 113

  19. [27]

    H., et al

    Helmi, A., Babusiaux, C., Koppelman, H. H., et al. 2018, Nature, 563, 85

  20. [28]

    F., Wetzel, A., Kere s , D., et al

    Hopkins, P. F., Wetzel, A., Kere s , D., et al. 2018, Monthly Notices of the Royal Astronomical Society, 480, 800

  21. [29]

    A., Gilmore , G., & Irwin , M

    Ibata , R. A., Gilmore , G., & Irwin , M. J. 1994, , 370, 194, 10.1038/370194a0

  22. [30]

    Jeans , J. H. 1915, , 76, 70, 10.1093/mnras/76.2.70

  23. [31]

    2022 a , , 510, 4706, 10.1093/mnras/stab3711

    Li , C., & Binney , J. 2022 a , , 510, 4706, 10.1093/mnras/stab3711

  24. [32]

    2022 b , , 516, 3454, 10.1093/mnras/stac1788

    ---. 2022 b , , 516, 3454, 10.1093/mnras/stac1788

  25. [33]

    2023, Monthly Notices of the Royal Astronomical Society, 524, 6331

    Li, C., Siebert, A., Monari, G., Famaey, B., & Rozier, S. 2023, Monthly Notices of the Royal Astronomical Society, 524, 6331

  26. [34]

    2024 a , , 690, A26, 10.1051/0004-6361/202449742

    Li , C., Yuan , Z., Monari , G., et al. 2024 a , , 690, A26, 10.1051/0004-6361/202449742

  27. [35]

    2024 b , , 527, 9767, 10.1093/mnras/stad3817

    Li , Z., Zhao , G., Zhang , R., et al. 2024 b , , 527, 9767, 10.1093/mnras/stad3817

  28. [36]

    Lin , D. N. C., & Tremaine , S. 1983, , 264, 364, 10.1086/160604

  29. [37]

    2021, Astronomy & Astrophysics, 649, A4

    Lindegren, L., Bastian, U., Biermann, M., et al. 2021, Astronomy & Astrophysics, 649, A4

  30. [38]

    McMillan, P. J. 2017, Monthly Notices of the Royal Astronomical Society, 465, 76

  31. [39]

    C., et al

    Minchev , I., Famaey , B., Quillen , A. C., et al. 2012, , 548, A126, 10.1051/0004-6361/201219198

  32. [40]

    C., & White , S

    Mo , H., van den Bosch , F. C., & White , S. 2010, Galaxy Formation and Evolution

  33. [41]

    2018, The Astrophysical Journal Letters, 863, L28

    Myeong, G., Evans, N., Belokurov, V., Sanders, J., & Koposov, S. 2018, The Astrophysical Journal Letters, 863, L28

  34. [42]

    C., Vasiliev , E., Iorio , G., Evans , N

    Myeong , G. C., Vasiliev , E., Iorio , G., Evans , N. W., & Belokurov , V. 2019, , 488, 1235, 10.1093/mnras/stz1770

  35. [43]

    2024, , 966, 108, 10.3847/1538-4357/ad35ba

    Nguyen , T., Ou , X., Panithanpaisal , N., et al. 2024, , 966, 108, 10.3847/1538-4357/ad35ba

  36. [44]

    2017, , 465, 1621, 10.1093/mnras/stw2819

    Portail , M., Gerhard , O., Wegg , C., & Ness , M. 2017, , 465, 1621, 10.1093/mnras/stw2819

  37. [45]

    E., Wetzel, A., Loebman, S., et al

    Sanderson, R. E., Wetzel, A., Loebman, S., et al. 2020, The Astrophysical Journal Supplement Series, 246, 6

  38. [46]

    2010, Monthly Notices of the Royal Astronomical Society, 403, 1829

    Sch \"o nrich, R., Binney, J., & Dehnen, W. 2010, Monthly Notices of the Royal Astronomical Society, 403, 1829

  39. [47]

    Sellwood , J. A. 2014, Reviews of Modern Physics, 86, 1, 10.1103/RevModPhys.86.1

  40. [48]

    2016, , 819, 92, 10.3847/0004-637X/819/2/92

    ---. 2016, , 819, 92, 10.3847/0004-637X/819/2/92

  41. [49]

    A., & Binney , J

    Sellwood , J. A., & Binney , J. J. 2002, , 336, 785, 10.1046/j.1365-8711.2002.05806.x

  42. [50]

    C., Gerhard , O., Portail , M., Vasiliev , E., & Clarke , J

    Sormani , M. C., Gerhard , O., Portail , M., Vasiliev , E., & Clarke , J. 2022, , 514, L1, 10.1093/mnrasl/slac046

  43. [51]

    Tremaine , S., & Weinberg , M. D. 1984, , 209, 729, 10.1093/mnras/209.4.729

  44. [52]

    2019, , 482, 1525, 10.1093/mnras/sty2672

    Vasiliev , E. 2019, , 482, 1525, 10.1093/mnras/sty2672

  45. [53]

    2019, , 485, 3296, 10.1093/mnras/stz572

    Wegg , C., Gerhard , O., & Bieth , M. 2019, , 485, 3296, 10.1093/mnras/stz572

  46. [54]

    2015, , 450, 4050, 10.1093/mnras/stv745

    Wegg , C., Gerhard , O., & Portail , M. 2015, , 450, 4050, 10.1093/mnras/stv745

  47. [55]

    Weinberg , M. D. 1986, , 300, 93, 10.1086/163785

  48. [56]

    1989, , 239, 549, 10.1093/mnras/239.2.549

    ---. 1989, , 239, 549, 10.1093/mnras/239.2.549

  49. [57]

    Widrow , L. M. 2023, , 522, 477, 10.1093/mnras/stad973

  50. [58]

    F., et al

    Yuan , Z., Li , C., Martin , N. F., et al. 2024, , 691, L1, 10.1051/0004-6361/202348593

  51. [59]

    W., Kane , S

    Zhang , H., Belokurov , V., Evans , N. W., Kane , S. G., & Sanders , J. L. 2024, , 533, 3395, 10.1093/mnras/stae2023

  52. [60]

    W., et al

    Zhang , H., Belokurov , V., Evans , N. W., et al. 2025, , 983, L10, 10.3847/2041-8213/adc261

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.