Pith. sign in

REVIEW 3 major objections 4 minor 36 references

A Lower Limit on the Mass of Our Galaxy from the H3 Survey

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The Milky Way weighs at least 0.91 trillion solar masses.

desk verdict A clean, transparent timing-argument lower limit on the Milky Way's mass; the headline number is plausible but the simulation calibration is thinner than the 90% confidence wording suggests. read the letter →

arxiv 1909.02025 v1 pith:5FORYPEN submitted 2019-09-04 astro-ph.GA

Dennis Zaritsky , Charlie Conroy , Huanian Zhang , Ana Bonaca , Nelson Caldwell , Phillip A. Cargile , Benjamin D. Johnson , Rohan P. Naidu This is my paper · ORCID
classification astro-ph.GA
keywords MilkyWaymasstimingargumentdarkmatterhaloouterstarsH3Surveylowerlimitgalacticdynamicsvirial
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 seeks to set a firm lower bound on the total mass of the Milky Way's dark-matter halo, a quantity that has resisted a precise measurement for decades. Using the timing argument—the idea that the most distant stars are still falling into the Galaxy after the initial expansion of the universe—it analyzes 32 stars beyond 60 kiloparsecs from the Galactic center in the H3 spectroscopic survey. Mock catalogs drawn from four Auriga simulations calibrate how much the simple timing argument underestimates the true halo mass, yielding a conservative correction factor. The result is that $M_{200}$, the mass enclosed within the radius where the mean density is 200 times the cosmic critical density, exceeds $0.91\times 10^{12}\,M_\odot$ with 90% confidence, with a preferred value near $1.4\times 10^{12}\,M_\odot$. If the limit is right, it excludes about half of the mass range allowed by several published Milky Way mass estimates and gives complex dynamical models a much-needed prior.

What carries the argument

The central object is the timing argument in the analytic form of Sandage (1986): for a tracer on a radial orbit around a point mass, the age of the universe fixes the orbital time, so the observed distance and radial velocity determine the enclosed mass. Because the simple argument systematically underestimates the true mass, the paper calibrates it with mock H3 catalogs built from four Auriga simulations; the maximum ratio $M_{\rm Timing}/M_{200}$ seen in any mock is 0.43, which sets the factor 2.3 used for the single most extreme star. A Kolmogorov-Smirnov comparison of the full distribution of $M_{\rm Timing}/M_{200}$ between the 32 observed stars and the mock catalogs supplies the second, distribution-based limit.

What would settle it

Measure accurate proper motions for the 32 outer halo stars and re-derive their full orbits: if the most constraining star, H3 117408280, proves to have substantial tangential velocity, the radial-orbit timing-argument model is violated and the $0.91\times 10^{12}\,M_\odot$ limit must be recomputed, and any newly found outer halo star with a timing mass above $0.49\times 10^{12}\,M_\odot$ would raise the limit.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is a conservative, model-calibrated lower limit on the Milky Way's halo mass, computed from a classic dynamical tool. The most extreme outer halo star in the H3 sample has a timing-argument mass of $0.49\times 10^{12}\,M_\odot$; in four Auriga simulations, no tracer ever returns a timing mass above 43% of the true $M_{200}$, so multiplying by $1/0.43\approx 2.3$ gives a single-star estimate of $1.1\times 10^{12}\,M_\odot$. A second, distribution-based route compares the full set of 32 scaled timing masses to the mock catalogs with a Kolmogorov-Smirnov test, rejecting $M_{200}$ values below $0.91\times 10^{12}\,M_\odot$ at 90% confidence and values above $2.13\times 10^{12}\,M_\odot$. The two routes agree, and the paper adopts the smaller limit as its headline result, while noting a preferred value near $1.4\times 10^{12}\,M_\odot$.

Load-bearing premise

The result rests on the assumption that the four Auriga simulations are a fair stand-in for the Milky Way's outer halo in both the mix of stellar orbits and the underlying gravitational potential, so the correction factor of 2.3 and the mock distribution of timing masses apply to the real Galaxy.

Editorial extensions

If this is right

  • Adopted as a prior, the limit cuts roughly half of the mass range allowed by many recent Milky Way mass measurements, including estimates based on the Sagittarius dwarf stream.
  • The agreement between the single-extreme-star route and the full-distribution route means the lower limit does not stand on one object alone.
  • The preferred mass near $1.4\times 10^{12}\,M_\odot$ is suggestive but not established; confirming it needs a larger sample of outer halo stars and a more complete treatment of the inner halo.
  • As planned surveys add more stars beyond 60 kiloparsecs, the same calibrated procedure should produce a stronger limit, because a larger sample samples the upper envelope of timing masses more fully.

Reading between the lines

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

  • If proper motions for these 32 stars become available, the inferred timing masses should rise, not fall, because non-radial orbital motion only makes the point-mass timing argument underestimate the true mass; the 90% lower limit would likely strengthen.
  • Applying the same mock-calibration protocol to other cosmological hydrodynamical simulations would test whether the factor 2.3 and the mock ratio distribution are particular to the Auriga runs or a general property of cold-dark-matter halos.
  • The same method could be exported to other nearby galaxies with resolved stellar halos, yielding a model-light lower mass limit outside the Local Group.
  • The one high-velocity star rejected from the 32-star sample deserves follow-up spectroscopy and binarity checks; if it is a genuine bound outer-halo member, it could become the most constraining tracer and push the limit higher.
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

3 major / 4 minor

Summary. This paper derives a lower limit on the Milky Way's M200 using the timing argument applied to 32 outer halo stars (R > 60 kpc) from the H3 Survey. The analysis proceeds in two steps: a single-star calibration uses the maximum M_Timing/M200 ratio (0.43) found in mock catalogs from four Auriga simulations to define a correction factor of 2.3, and a complementary Kolmogorov-Smirnov test compares the distribution of M_Timing/M200 for the H3 stars, rescaled by trial M200 values, to the same mock distribution. The paper's headline result is M200 > 0.91 x 10^12 Msun at 90% confidence, with a preferred value near 1.4 x 10^12 Msun, and it argues that this limit should be used as a prior in more complex mass-modeling analyses.

Significance. If the result holds, this is a useful and non-trivial prior for Milky Way mass modeling, and it has the virtue of being simple, transparent, and based on distant tracers where extrapolation in radius is modest. The use of external, published simulations for calibration avoids circularity with the measured Milky Way mass, and the agreement between the extreme-star method and the distribution-based method is reassuring. The main significance risk is that the calibration rests on a small number of simulated halos, so the central methodological contribution is only as strong as the representativeness of those runs.

major comments (3)
  1. [Section 3.1] The calibration factor 2.3 = 1/0.43 is derived from the maximum M_Timing/M200 among 257 mock tracer particles in only four Auriga runs, after runs 16 and 21 were excluded by visual comparison of large-R substructure. Because both the single-star limit and the KS limit use this same calibration, the two quoted 90% limits are not independent with respect to the dominant systematic, namely the assumption that these four halos faithfully represent the Milky Way's outer-halo phase-space distribution and potential. The paper explicitly acknowledges this in Section 3.2, but the confidence statement is still conditional on that assumption. I request a quantitative sensitivity analysis: for example, increasing f_max from 0.43 to 0.50 would lower the single-star 90% limit from 0.97 to roughly 0.84 x 10^12 Msun, below the headline 0.91 x 10^12 Msun. Reporting the limits obtained from each Auriga run separately, and perhaps from all six runs including 16 and 21, would show how much of the 90% confidence is driven by the four-run selection.
  2. [Section 3.2] The KS test pools the four Auriga simulations into a single reference distribution and treats it as a deterministic model. With only four independent simulation halos, the >90% confidence statement does not include the sampling variance of the halo-to-halo distribution of M_Timing/M200. A bootstrap over the four halos, or a hierarchical treatment that regards each simulation as one draw from a population of possible Milky Way-like halos, is needed to show that the lower limit of 0.91 x 10^12 Msun is robust to which simulations are used. As written, the test can only reject a trial M200 relative to this particular four-run library, not relative to the population of possible halo phase-space distributions.
  3. [Section 2 and Figure 1] The sample definition excludes the one R > 60 kpc star with v_GSR_R < -500 km/s on the grounds that it is either a bad parameter fit or a physically compelling outlier, with an unresolved binary suggested as a possible cause. This exclusion is made before the mass analysis and is not justified quantitatively. Since the paper's distribution-based argument depends on the outer halo star sample, the authors should either provide a more robust justification for removing this star or repeat the analysis with it included to demonstrate that the quoted lower limit is unchanged.
minor comments (4)
  1. [Section 2] The paper uses a cut on tangential velocity v_T < 1000 km/s and a cut on GSR radial velocity, but it does not state how proper motions enter the GSR radial velocity for these distant stars; a sentence explaining the coordinate transformation and the role of proper-motion uncertainties would help the reader assess the 400 km/s selection.
  2. [Figure 3] The caption says 'inbound and outbound stars by closed and open symbols,' but these symbols are not defined in a legend in the printed figure; please add a legend or define them explicitly in the caption.
  3. [References] The text cites Grand et al. (2019) in the introduction, and the reference list contains both Grand et al. (2017) and Grand et al. (2019); these should be disambiguated clearly in all citation callouts.
  4. [Appendix A] The polynomial fit for R200 versus M200 is quoted with more significant digits than the data warrant; report the fit with uncertainties or show residual scatter in the figure.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the quoted lower limit is calibrated on external Auriga simulations and H3 data, not on the Milky Way mass being inferred.

full rationale

The central mass limit M200 > 0.91e12 Msun is not equivalent to any input by construction. Timing-argument masses are computed from the Sandage (1986) equations using H3 distances and radial velocities; the calibration factor 2.3 and the KS-test reference distribution are derived from the published Auriga simulations of Grand et al. (2017), which provide their own independent M200 and R200 values. The Milky Way's M200 is not used as an input to this calibration. The R200-M200 relation in Appendix A is fitted to the same external simulations, but it is used only for rescaling apocenters for presentation and for the KS comparison; the quoted limit does not reduce to this relation. The self-citations to Conroy et al. (2019) and Cargile et al. (2019) are references to the H3 survey data products, which are a data source rather than a load-bearing theoretical premise. The paper explicitly acknowledges the model-dependence of the KS approach in Section 3.2 ('we are assuming that the model is a fair representation of both the tracer distribution in phase space and the underlying potential'), but this is a stated systematic assumption, not a circular derivation. The conclusions may be sensitive to the choice of the four Auriga halos and to the finite mock sample, but such sensitivity is a robustness concern, not circularity.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The analysis depends on one fitted calibration factor (2.3) derived from the maximum ratio in four Auriga models, an empirical M200-R200 relation used for plotting, and a hand-chosen radial velocity cut of 400 km/s. The timing argument itself rests on standard assumptions about radial orbits, point masses, and the Hubble flow. No new physical entities are introduced.

free parameters (3)
  • Timing-to-M200 calibration factor C = 2.3 (inverse of 0.43 maximum ratio in four Auriga models)
    Used to convert the largest H3 timing mass into a lower limit on M200 (Section 3.1). It is a single number derived from the most extreme mock particles and carries statistical uncertainty that is not propagated.
  • R200-M200 polynomial coefficients = 128.968, 95.7676, -13.8829
    Empirical fit to the Auriga models used to convert assumed M200 to R200 when rescaling apocenters for visual comparison (Appendix A). Not central to the adopted lower limit, but used in Figure 3.
  • Radial velocity selection bound = 400 km/s
    The box |v_GSR_R| < 400 km/s is chosen to exclude one outlier star at v < -500 km/s; this hand-chosen cut affects the sample composition and hence both mass estimates (Section 2).
assumptions (5)
  • domain assumption The universe expands with the Hubble flow and tracer particles initially move outward with it, as assumed in the timing argument.
    Invoked in Section 1 and 3 (Sandage 1986). If outer halo stars did not start near the Hubble flow, the timing mass equations do not describe their orbits.
  • domain assumption The Milky Way is treated as a point mass and star orbits are radial (no tangential velocity).
    The timing argument equations assume this; the authors note that non-zero tangential velocity makes the inferred mass a lower limit, and rely on simulations to calibrate the bias (Section 1).
  • domain assumption The four selected Auriga halos are representative of the Milky Way's outer halo phase-space structure and potential.
    Used to calibrate the 2.3 factor and to build the comparison distribution for the KS test (Section 3.1-3.2). The paper explicitly acknowledges this assumption's fragility.
  • domain assumption The H3 survey distances and stellar parameters are unbiased to within quoted uncertainties.
    Distances dominate the timing mass errors; they rely on the pipeline of Cargile et al. (2019) and Conroy et al. (2019).
  • domain assumption The adopted age of the universe, 13.75 Gyr, is correct.
    The orbital time is set to this value (Section 3), so a different cosmic age changes the inferred mass.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Lower Limit on the Mass of Our Galaxy from the H3 Survey." pith.science (2026). https://pith.science/paper/5FORYPEN

@misc{pith2026190902025,
  author       = {Pith},
  title        = {Pith review of: A Lower Limit on the Mass of Our Galaxy from the H3 Survey},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5FORYPEN}},
  note         = {Machine review of arXiv:1909.02025}
}
abstract

The timing argument provides a lower limit on the mass of the Milky Way. We find, using a sample of 32 stars at $R > 60$ kpc drawn from the H3 Spectroscopic Survey and mock catalogs created from published numerical simulations, that M$_{200} > 0.91\times 10^{12}$ M$_\odot$ with 90% confidence. We recommend using this limit to refine the allowed prior mass range in more complex and sophisticated statistical treatments of Milky Way dynamics. The use of such a prior would have significantly reduced many previously published uncertainty ranges. Our analysis suggests that the most likely value of M$_{200}$ is $\sim 1.4 \times 10^{12}$ M$_\odot$, but establishing this as the Milky Way mass requires a larger sample of outer halo stars and a more complete analysis of the inner halo stars in H3. The imminent growth in the sample of outer halo stars due to ongoing and planned surveys will make this possible.

Figures

Figures reproduced from arXiv: 1909.02025 by the authors.

Figure 1
Figure 1. The H3 sample and outer halo stars. From the full sample we select a sample of 32 outer halo stars (R > 60 kpc, −400 ≤ v GSR R ≤ 400 km sec−1 ) for the timing argument analysis, as shown highlighted within the dotted selection box. The basic timing argument is simple in its application. Using the equations provided by Sandage (1986), we find the smallest orbital phase, θ, that solves the equations relating distance,… view at source ↗
Figure 2
Figure 2. Mock H3 catalogs created from Auriga models (6, 23, 24, and 27) from Grand et al. (2017). The outer halo star selection box from [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Distribution of scaled timing argument results. We scale the derived mass by M200 for each of the outer halo star particles in each of the four simulations and the calculated apocenter distances by R200 in the upper panel. The different models are represented by different colors and inbound and outbound stars by closed and open symbols, respectively. All estimates are less than half of the true value regardless of t… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: A sample of literature measurement of the Milky Way mass. The shaded region illustrates the region excluded by our 90% confidence lower limit. Only two previous mea￾surements are strongly argued against, but for most of the measurement our lower limit reduces the allow…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 26 canonical work pages

  1. [1]

    F., McConnachie, A

    Barber, C., Starkenburg, E., Navarro, J. F., McConnachie, A. W., & Fattahi, A. 2014, MNRAS, 437, 959

    work page 2014

  2. [2]

    1985, ApJS, 58, 39

    Bertschinger, E. 1985, ApJS, 58, 39

    1985

  3. [3]

    S., Sohn, S

    Boylan-Kolchin, M., Bullock, J. S., Sohn, S. T., Besla, G., & van der Marel, R. P. 2013, ApJ, 768, 140

    work page 2013

  4. [4]

    R., Geller, M

    Brown, W. R., Geller, M. J., & Kenyon, S. J. 2014, ApJ, 787, 89

    2014

  5. [5]

    MINESweeper: Spectrophotometric Modeling of Stars in the Gaia Era

    Cargile, P. A., Conroy, C., Johnson, B. D., et al. 2019, arXiv e-prints, arXiv:1907.07690

    work page Pith review arXiv 2019

  6. [6]

    G., & Gottlöber, S

    Carlesi, E., Hoffman, Y., Sorce, J. G., & Gottlöber, S. 2017, MNRAS, 465, 4886

    work page 2017

  7. [7]

    S., van de Weygaert, R., Hellwing, W

    Cautun, M., Frenk, C. S., van de Weygaert, R., Hellwing, W. A., & Jones, B. J. T. 2014, MNRAS, 445, 2049

    work page 2014

  8. [8]

    Mapping the Stellar Halo with the H3 Spectroscopic Survey

    Conroy, C., Bonaca, A., Cargile, P., et al. 2019, arXiv e-prints, arXiv:1907.07684

    work page Pith review arXiv 2019

Show all 36 references
  1. [9]

    C., Abrams, D

    Dalton, G., Trager, S. C., Abrams, D. C., et al. 2012, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 8446, Ground-based and Airborne Instrumentation for Astronomy IV, 84460P 8 Zaritsky et al. Figure A.1. R200 vs. M200 for simulated halos ...

    2012

  2. [10]

    J., Fattahi, A., Belokurov, V., et al

    Deason, A. J., Fattahi, A., Belokurov, V., et al. 2019, MNRAS, 485, 3514

    2019

  3. [11]

    J., Liu, C., et al

    Deng, L.-C., Newberg, H. J., Liu, C., et al. 2012, Research in Astronomy and Astrophysics, 12, 735

    2012

  4. [12]

    V., & More, S

    Diemer, B., Mansfield, P., Kravtsov, A. V., & More, S. 2017, ApJ, 843, 140

    2017

  5. [13]

    Dierickx, M. I. P., & Loeb, A. 2017, ApJ, 847, 42

    2017

  6. [14]

    A., van der Marel, R

    Fardal, M. A., van der Marel, R. P., Law, D. R., et al. 2019, MNRAS, 483, 4724

    2019

  7. [15]

    A., & Goldreich, P

    Fillmore, J. A., & Goldreich, P. 1984, ApJ, 281, 1

    1984

  8. [16]

    2017, New Astronomy, 55, 32

    Fragione, G., & Loeb, A. 2017, New Astronomy, 55, 32

    2017

  9. [17]

    Gibbons, S. L. J., Belokurov, V., & Evans, N. W. 2014, MNRAS, 445, 3788

    2014

  10. [18]

    Grand, R. J. J., Deason, A. J., White, S. D. M., et al. 2019, arXiv e-prints, arXiv:1905.09834

    2019 arXiv

  11. [19]

    Grand, R. J. J., Gómez, F. A., Marinacci, F., et al. 2017, MNRAS, 467, 179

    2017

  12. [20]

    Guiglion, G., Battistini, C., Bell, C. P. M., et al. 2019, The Messenger, 175, 17

    2019

  13. [21]

    2013, ApJS, 208, 19 Kafle, P

    Hinshaw, G., Larson, D., Komatsu, E., et al. 2013, ApJS, 208, 19 Kafle, P. R., Sharma, S., Lewis, G. F., & Bland-Hawthorn, J. 2014, ApJ, 794, 59

    2013

  14. [22]

    D., & Woltjer, L

    Kahn, F. D., & Woltjer, L. 1959, ApJ, 130, 705

    1959

  15. [23]

    L., & Carlberg, R

    Kroeker, T. L., & Carlberg, R. G. 1991, ApJ, 376, 1

    1991

  16. [24]

    Li, Y.-S., & White, S. D. M. 2008, MNRAS, 384, 1459

    2008

  17. [25]

    R., Schiavon, R

    Majewski, S. R., Schiavon, R. P., Frinchaboy, P. M., et al. 2017, AJ, 154, 94

    2017

  18. [26]

    L., Sharma, S., Buder, S., et al

    Martell, S. L., Sharma, S., Buder, S., et al. 2017, MNRAS, 465, 3203

    2017

  19. [27]

    W., & Walker, M

    Mateo, M., Olszewski, E. W., & Walker, M. G. 2008, ApJ, 675, 201

    2008

  20. [28]

    2017, MNRAS, 468, 3428

    Patel, E., Besla, G., & Mandel, K. 2017, MNRAS, 468, 3428

    2017

  21. [29]

    Patel, E., Besla, G., Mandel, K., & Sohn, S. T. 2018, ApJ, 857, 78 Piffl, T., Scannapieco, C., Binney, J., et al. 2014, A&A, 562, A91

    2018

  22. [30]

    V., Navarro, J

    Sales, L. V., Navarro, J. F., Abadi, M. G., & Steinmetz, M. 2007, MNRAS, 379, 1475

    2007

  23. [31]

    L., Evans, N

    Watkins, L. L., Evans, N. W., & An, J. H. 2010, MNRAS, 406, 264

    2010

  24. [32]

    White, S. D. M., & Zaritsky, D. 1992, ApJ, 394, 1

    1992

  25. [33]

    A., Belokurov, V., Casey, A

    Williams, A. A., Belokurov, V., Casey, A. R., & Evans, N. W. 2017, MNRAS, 468, 2359

    2017

  26. [34]

    X., Rix, H

    Xue, X. X., Rix, H. W., Zhao, G., et al. 2008, ApJ, 684, 1143

    2008

  27. [35]

    2017, MNRAS, 465, 3724

    Zaritsky, D., & Courtois, H. 2017, MNRAS, 465, 3724

    2017

  28. [36]

    W., Schommer, R

    Zaritsky, D., Olszewski, E. W., Schommer, R. A., Peterson, R. C., & Aaronson, M. 1989, ApJ, 345, 759

    1989

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.