Kinematic distances of galaxies in the Local Volume

A.A.Popova; I.D.Karachentsev

REVIEW 3 major objections 4 minor 1 cited by

Kinematic distances of galaxies in the Local Volume

T0 review · 3 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Kinematic distance estimates for Local Volume galaxies hold up at 20% accuracy, except near Virgo and Coma I.

desk verdict Useful comparison, but the 20% accuracy claim is probably measuring the NAM model in-sample against the very TRGB distances used to build it. read the letter →

arxiv 2412.03988 v1 pith:7RLDFF6F submitted 2024-12-05 astro-ph.GA

I.D.Karachentsev , A.A.Popova This is my paper · ORCID

classification astro-ph.GA PACS 98.62.Py98.65.-r95.85.-e

keywords NumericalActionMethodkinematicdistancesTipoftheRedGiantBranchLocalVolumepeculiarvelocityfieldgalaxyVirgoclusterCosmicflows

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper asks whether the Numerical Action Method (NAM), which derives galaxy distances from measured radial velocities plus a model of local gravity, can replace the expensive Tip of the Red Giant Branch (TRGB) technique for nearby galaxies. Comparing 418 TRGB distances from the Local Volume catalog with the NAM kinematic distances, the authors find a small scale offset of -0.30 Mpc and a scatter of 1.57 Mpc, implying about 20% accuracy per galaxy and 15% for group members. They conclude that NAM distances are dependable for roughly 90% of the sky, with the notable exceptions being the Virgo cluster infall zone and the Coma I group. If correct, this makes accurate distance and luminosity estimates for thousands of nearby galaxies cheap and fast, limited mainly by the availability of radial velocity measurements.

What carries the argument

The Numerical Action Method (NAM) is the central mechanism: it reconstructs galaxy orbits and distances by modeling the local peculiar velocity field, using the positions and masses of nearby galaxy groups as attractors, including the influence of the Virgo cluster and the Local Void. The comparison benchmark is the Tip of the Red Giant Branch (TRGB) distance scale, whose ~5% per-galaxy accuracy comes from HST photometry. The paper relies on the published NAM model of Kourkchi et al. 2020, reading off NAM distances from its diagram/calculator, and compares them statistically (mean offset, standard deviation, Gaussian fit) against the TRGB distances in the Local Volume catalog.

What would settle it

Recompute the NAM model of Kourkchi et al. 2020 while withholding the 418 TRGB distances from its input constraints, then measure the scatter between those withheld TRGB distances and the new model's predicted distances; if the scatter exceeds ~20% (or the mean offset changes substantially), the paper's claimed predictive accuracy is overstated.

Watch

Extended reading notes

Core claim

Using 418 galaxies with TRGB distances and excluding 12 strong outliers, the paper finds that NAM-derived distances agree with TRGB distances with a mean difference of -0.30 ± 0.08 Mpc and a standard deviation of 1.57 Mpc. After quadratically subtracting the 5% TRGB error, the NAM distance error is about 19% for individual galaxies, improving to about 15% when the average is taken over the members of a populated group. The systematic offset is consistent with comparisons against group-membership distances and Tully-Fisher distances, and the main failures are confined to the Virgo cluster infall region and the Coma I group, covering about 10% of the sky.

Load-bearing premise

The 418 TRGB distances used as the benchmark were not used as constraints when the NAM model of Kourkchi et al. 2020 was constructed, so the measured 20% scatter is a genuinely independent test rather than an in-sample fit.

Editorial extensions

If this is right

For most galaxies in the Local Volume, distances can be estimated from radial velocities and the NAM model with roughly 20% accuracy, without costly HST observations.
Within galaxy groups containing several measured radial velocities, averaging NAM distances improves the typical error to about 15%.
NAM distances are competitive with or better than Tully-Fisher distances for the Local Volume population, where the paper estimates a 31% Tully-Fisher error.
The known local velocity anomalies (Virgo infall and the Coma I group with its anomalous peculiar velocities, plus the Leo Spur region) must be avoided for NAM distance work, constraining the method's reliable sky footprint to about 90%.
The small systematic offset of -0.30 Mpc implies a mild difference in the distance scale between the NAM model and the TRGB scale, which could affect volume-limited samples of nearby galaxies.

Reading between the lines

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

A testable extension would be to split the comparison by projected angular distance from the Virgo cluster center and from the Coma I group, to map exactly where NAM scatter exceeds its nominal 20% and where it remains valid at the group-average 15% level.
The claimed accuracy naturally extrapolates to using 21-cm HI velocity surveys (e.g., future wide-area radio surveys) as a cheap pipeline for distances and luminosities of dwarf galaxies, but only outside the two anomalous regions.
One could directly test the scale offset by comparing NAM distances against a larger sample of Cepheid or SN Ia distances that were not used in constructing the NAM model.
The method's demonstrated failure in the Coma I group and the Virgo infall zone suggests that any kinematic reconstruction of local velocities must explicitly model infall onto Virgo and the anomalous negative-velocity flow north of it before its distances can be trusted there.

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.

Desk Editor's Note

read the letter

Key point: the paper compares NAM kinematic distances with 418 HST TRGB distances, finds a 1.57 Mpc scatter, and declares 20% accuracy. But the NAM model in [8] was built using about 400 TRGB distances, and the paper never shows the 418 benchmark galaxies are not the same ones. So the quoted scatter is probably how well the model reproduces its own input distances, not how well it predicts distances for new galaxies. The 20% claim stands only after an out-of-sample test.

Credit where due: the comparison itself is new, the writing is transparent, and the outlier list is useful. The authors correctly identify Virgo and Coma I as regions where NAM fails, and the group-membership and Tully-Fisher comparisons are sensible supporting checks even if none of them is independent of the possible overlap. The 5% TRGB error subtraction and the quoted offsets are clearly explained.

Soft spots, in order. First, the circularity issue above is load-bearing. Section 1 says 'the main array consisted of high-precision TRGB distances ... about 400,' and Section 2 compares to 418 TRGB galaxies from the same LVGDB. Those numbers are too close to ignore. A short check of whether the 418 appear in the Cosmicflows-3 constraint set, or a split-sample validation, would settle it. Without that, the 20% is an upper bound on accuracy, not the claimed estimate.

Second, the 20% number comes after removing 12 of 430 galaxies (and 16 of 159 TF galaxies). Removing outliers is defensible if done transparently, and it is here, but the excluded galaxies cluster in Virgo and Coma I. Since those are also the regions where NAM is least constrained, treating them as an external caveat rather than part of the test makes the '90% of the sky' claim optimistic.

Third, the systematic offset = -0.30 Mpc is reported but not modeled. At 8 Mpc mean distance it is a few percent, so minor for most purposes, but it hints at a scale problem in NAM that the paper does not address.

Fourth, the machine-readable table is only available on request. For a validation paper that is trying to convince people to use NAM distances, that is an unnecessary barrier.

Who is this for? People who need quick distances for many Local Volume galaxies and cannot get HST time. They will want to read this, but they should not adopt the 20% figure until the overlap question is resolved. I would send it to a referee; the issue is important and fixable. If the authors demonstrate out-of-sample behavior, this becomes a solid reference. As it stands, treat the 20% claim as provisional.

Referee Report

3 major / 4 minor

Summary. The paper compares kinematic distances from the Numerical Action Method (NAM) model of Cosmicflows-3 (Kourkchi et al. 2020) with high-precision TRGB distances from the Local Volume Galaxy Database for 430 galaxies, and also with group-membership and Tully-Fisher distances. After excluding a small number of outliers identified with the Virgo and Coma I regions, the authors report a mean difference <D_NAM - D_TRGB> = -0.30 ± 0.08 Mpc, a standard deviation of 1.57 Mpc, and a relative error of about 19% after subtracting a 5% TRGB error in quadrature. They conclude that NAM distances are accurate to ~20% for individual galaxies over ~90% of the sky, and to ~15% when group-averaged velocities are used, except near Virgo and Coma I.

Significance. If established, the claim is practically valuable: NAM distances, derived from coordinates and radial velocities alone, could substitute for much more expensive HST TRGB distances for many nearby galaxies, enabling large statistical samples in the Local Volume. The paper is concise, uses a large homogeneous TRGB comparison sample, provides a transparent list of outliers, and cross-checks with two independent distance indicators. The main result is quantitative and directly falsifiable, and the underlying analysis is straightforward. However, the significance hinges on whether the comparison is genuinely out-of-sample with respect to the NAM model construction.

major comments (3)

[Section 1 and Section 3] The out-of-sample status of the TRGB comparison sample is not established. The Introduction states that the NAM model of Kourkchi et al. [8] was constructed using about 400 TRGB distances as the main constraint array, and the paper then validates that model against 418 TRGB distances from the LVGDB. The paper never demonstrates that these 418 galaxies are not the same objects used in constructing the model. If the two sets overlap, the scatter of 1.57 Mpc and the resulting 20% accuracy estimate are in-sample fit statistics, not predictive accuracies. The authors must quantify the overlap between the LVGDB TRGB sample and the distance catalog used by Kourkchi et al. [8], and if the overlap is non-empty, recompute the scatter for the subset of galaxies that were not used as NAM constraints, or otherwise show that excluding them does not change the conclusion.
[Section 3, Fig. 2, and Table 1] The headline scatter sigma_delta = 1.57 Mpc is derived after excluding 12 of 430 galaxies with |Delta| > 5 Mpc, and the claim of ~90% sky coverage depends on this exclusion. The threshold is post-hoc, and the excluded objects are concentrated in the Virgo infall zone and the Coma I group, which are then described as regions where NAM is inapplicable. To make the accuracy claim robust, the authors should report the scatter for the full sample as well as for the cleaned sample, and should justify the threshold or treat these regions as separate regimes rather than as outliers. As written, the analysis removes exactly the problematic parts of the sky before computing the statistic that is then generalized to 90% of the sky, which risks overstating the reliability.
[Section 5] The systematic offset <D_NAM - D_TRGB> = -0.30 ± 0.08 Mpc is statistically significant but is not incorporated into the stated accuracy. The paper quotes a relative error of ~19% based on the scatter after subtracting the 5% TRGB error in quadrature, but the offset represents a zero-point bias of about 4% at the mean distance of 8.0 Mpc. A complete accuracy statement should either combine random and systematic terms (e.g., root-mean-square about zero) or provide a zero-point correction and test its stability. Without this, the claim that NAM distances have an accuracy of 20% is not fully quantified.

minor comments (4)

[Abstract] The abstract says the TRGB comparison uses 418 distances, while Section 2 describes 430 galaxies and Section 3 excludes 12 outliers before the Gaussian fit. Please make explicit in the abstract that the 418 number refers to the cleaned sample after outlier removal.
[Section 3] The phrase 'according to the diagram [8]' is vague; please specify how the NAM distances were obtained (e.g., the Cosmicflows-3 NAM distance calculator) so that the procedure is reproducible.
[Section 4] The statement 'Assuming that the relative error of NAM distances is 15%, we find approximately the same value for the error of mem-distances, 15%' is circular: the 15% NAM error is itself estimated in Section 3, not independently measured. Please present the variance decomposition without assuming the NAM error, or state the assumption explicitly as a limitation.
[Table 2 and text] There are minor typographical issues: 'NG3115' should be 'NGC3115'; 'Cosmic-flow-3' should be 'Cosmicflows-3'; and the table heading 'LVGB' should be 'LVGDB'. Also, the text says '16 galaxies in the table with TRGB distances' but the table appears to contain a different count after adding the four non-TRGB measurements; please reconcile the count.

Circularity Check

1 steps flagged · score 6.0 of 10

NAM accuracy claim may be in-sample: the ~400 TRGB distances used to construct the [8] NAM model overlap the 418-galaxy TRGB validation sample, with no out-of-sample demonstration.

fitted input called prediction [Section 1 (construction of NAM model), Section 2 ('NAM-distances were determined according to the diagram [8]'), Section 3 (Figs. 1-2, the 19% error estimate), Abstract.]
"Kourkchi et al. [8] developed a convenient scheme for determining the kinematic (NAM) distances of LV galaxies from their coordinates and radial velocity Vg relative to the center of our Galaxy. In this case, estimates of the distance of galaxies made by various methods were used. The main array consisted of high-precision measurements of distances by the luminosity of the Tip of the Red Giant Branch (TRGB) with a total number of about 400. ... For the remaining 430 galaxies, kinematic NAM-distances were determined according to the diagram [8]."

The validation sample in the Abstract and Section 3 is 418 TRGB galaxies, while the NAM model [8] is stated in Section 1 to have been built using about 400 TRGB distances as input. The paper never shows that the 418 LVGDB benchmark galaxies are disjoint from those ~400 model inputs, so the comparison can be an in-sample check: the NAM distance field was constructed with those same TRGB distances and is then being measured against them. The reported <D_NAM - D_TRGB> = -0.30 +/- 0.08 Mpc and sigma_Delta = 1.57 Mpc (19% relative error) are therefore a goodness-of-fit statistic, not an out-of-sample predictive accuracy.

full rationale

The paper's central empirical claim is that NAM distances are accurate to ~20% within the Local Volume, based on comparing NAM distances from the Cosmicflows-3 model [8] with 418 TRGB distances from the LVGDB. The paper itself states that the NAM model was constructed using roughly 400 TRGB distances, and it does not demonstrate that the 418 validation galaxies are independent of those inputs. This makes the headline accuracy estimate potentially an in-sample fit statistic rather than a predictive validation. The NAM reconstruction is not a per-galaxy curve fit, so the result is not forced to equal the input by construction; thus a moderate score is appropriate. Other cited prior work (LVGDB, group membership, Tully-Fisher comparisons) is used as data or independent comparison and does not create additional circularity. No machine-checked or externally falsifiable out-of-sample test is provided. The proper remedy would be to exclude from the comparison any TRGB galaxy that entered the construction of the NAM model, or to validate against newly measured distances obtained after the model was built.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central 20% accuracy claim rests on adopted outlier thresholds and on the NAM model from prior work. No new entities are introduced. The most important unstated premise is that the comparison TRGB sample is independent of the NAM model's construction.

free parameters (4)

Outlier threshold |Δ| > 5 Mpc for TRGB-NAM comparison = 5 Mpc
12 of 430 galaxies with |Δ| > 5 Mpc are excluded before estimating the 1.57 Mpc scatter and the 20% accuracy claim (Section 3).
Membership-distance outlier threshold |Δ| > 5 Mpc = 5 Mpc
9 of 94 membership-distance galaxies are excluded before computing scatter (Section 4).
Tully-Fisher outlier threshold |Δ| > 16 Mpc = 16 Mpc
16 of 159 TF galaxies are excluded before computing scatter (Section 4).
Adopted TRGB relative error of 5% = 5%
Quoted from the literature and subtracted in quadrature from the observed scatter to claim a 19% NAM relative error (Section 3).

assumptions (5)

domain assumption NAM distances from the Cosmicflows-3 model (Kourkchi et al. 2020) are valid inputs; the paper takes them as given without re-deriving or validating the model.
Section 2 states NAM distances were determined according to the diagram from [8].
domain assumption Radial velocity errors are negligible (stated < 8 km/s, corresponding to < 0.1 Mpc in DNAM).
Section 2; if velocity errors were larger for some galaxies, the claimed accuracy would degrade.
domain assumption TRGB distances are accurate to about 5% and are normally distributed, so their error can be subtracted in quadrature from the scatter of differences.
Sections 1 and 3; no re-derivation of TRGB errors is given.
domain assumption The NAM and TRGB distance errors are independent and uncorrelated.
Used in the quadratic subtraction in Section 3; if the NAM model used the same TRGB distances as constraints, this assumption fails.
ad hoc to paper Galaxies with |Δ| > 5 Mpc (or > 16 Mpc for TF) can be excluded as belonging to known anomalous regions (Virgo, Coma I, erroneous TRGB), and their exclusion does not bias the accuracy estimate for the remaining sky.
Sections 3 and 4: exclusions are justified post hoc; the central 20% accuracy claim is estimated on the cleaned sample.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Kinematic distances of galaxies in the Local Volume." pith.science (2026). https://pith.science/paper/7RLDFF6F

@misc{pith2026241203988,
  author       = {Pith},
  title        = {Pith review of: Kinematic distances of galaxies in the Local Volume},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7RLDFF6F}},
  note         = {Machine review of arXiv:2412.03988}
}

read the original abstract

We consider the kinematic distances to nearby galaxies obtained by the Numerical Action Method (NAM) based on the Cosmic-flow-3 survey data. NAM distances are compared with 418 high-precision distances measured by the Tip of the Red Giant Branch (TRGB) method using the Hubble Space Telescope. We estimated the average difference <D_NAM - D_TRGB> = -0.30 +- 0.08 Mpc and the standard deviation of 1.57 Mpc. Approximately the same difference in the distance scale is obtained in comparison with less accurate distance estimates through the membership of galaxies in known groups or from the Tully-Fisher relation. We conclude that the NAM method provides distance estimates with an accuracy of 20% within the Local Volume, which is valid for ~90% of the sky, except for the regions of the Virgo cluster and the Coma-I group.

Figures

Figures reproduced from arXiv: 2412.03988 by the authors.

**Figure 2.** Distribution of number of galaxies in the Local Volume by the [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

New optical velocities for local galaxy candidates
astro-ph.GA 2026-07 accept novelty 5.0 of 10

BTA optical velocities confirm 12 new Local Volume companions (including satellites of DDO 46, NGC 2903, NGC 4826) and yield NGC 1068 group mass (2.6 ± 1.0) × 10^12 M⊙.

Reference graph

Works this paper leans on

16 extracted references · 16 canonical work pages · cited by 1 Pith paper

[8]

J., 159, 67, 2020

E.Kourkchi, H.M.Courtois, R.Graziani et al., Astron. J., 159, 67, 2020

work page 2020
[1]

Kraan-Korteweg, G.A.Tammann, Astron

R.C. Kraan-Korteweg, G.A.Tammann, Astron. Nachr., 300, 181, 1979

work page 1979
[2]

J., 127, 2031, 2004

I.D.Karachentsev, V.E.Karachentseva, W.K.Huchtmeier, D.I.Maka rov, Astron. J., 127, 2031, 2004

work page 2004
[3]

J., 145, 101, 2013

I.D.Karachentsev, D.I.Makarov, E.I.Kaisina, Astron. J., 145, 101, 2013

work page 2013
[4]

Bull., 67, 115, 2012

E.I.Kaisina, D.I.Makarov, I.D.Karachentsev, S.S.Kaisin, Astrophys . Bull., 67, 115, 2012

work page 2012
[5]

J., 676, 184, 2008

R.B.Tully, E.J.Shaya, I.D.Karachentsev et al., Astrophys. J., 676, 184, 2008

work page 2008
[6]

P. J. E.Peebles, S.D.Phelps, E.J.Shaya, R.B.Tully, Astrophys. J., 554, 104, 2001

work page 2001
[7]

J., 850, 207, 2017

E.J.Shaya, R.B.Tully, Y.Hoﬀman, D.Pomar´ ede, Astrophys. J., 850, 207, 2017

work page 2017

Show all 16 references

[9]

K.N.Abazajian, J.K.Adelman-McCarthy, M.A.Ag¨ ueros et al., Astrop hys. J. Suppl., 182, 543, 2009

2009
[10]

J., 128, 16, 2004

B.S.Koribalski, L.Staveley-Smith, V.A.Kilborn et al., Astron. J., 128, 16, 2004

2004
[11]

J., 142, 170, 2011

M.P.Haynes, R.Giovanelli, A.M.Martin et al., Astron. J., 142, 170, 2011

2011
[12]

C.-P.Zhang, M.Zhu, P.Jiang et al., Science China Physics, Mechanics , and Astronomy, 67, 219511, 2024

2024
[13]

J.Alonso-Garcia, M.Mateo, A.Aparicio, Publ. Astron. Soc. Pac., 118, 580, 2006

2006
[14]

J., 743, 123, 2011

I.D.Karachentsev, O.G.Nasonova, H.M.Courtois, Astrophys. J., 743, 123, 2011. 15.I.D. Karachentsev, E.I.Kaisina, V.E.Karachentseva, Mon. Not. R oy. Astron. Soc., 521, 840, 2023

2011
[16]

R.B.Tully, Nearby Galaxies Catalog (Cambridge: Cambridge Univers ity Press), 1988

1988
[17]

and Astrophys., 54, 661, 1977

R.B.Tully, J.R.Fisher, Astron. and Astrophys., 54, 661, 1977. 7

1977

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [8]

J., 159, 67, 2020

E.Kourkchi, H.M.Courtois, R.Graziani et al., Astron. J., 159, 67, 2020

work page 2020

[2] [1]

Kraan-Korteweg, G.A.Tammann, Astron

R.C. Kraan-Korteweg, G.A.Tammann, Astron. Nachr., 300, 181, 1979

work page 1979

[3] [2]

J., 127, 2031, 2004

I.D.Karachentsev, V.E.Karachentseva, W.K.Huchtmeier, D.I.Maka rov, Astron. J., 127, 2031, 2004

work page 2004

[4] [3]

J., 145, 101, 2013

I.D.Karachentsev, D.I.Makarov, E.I.Kaisina, Astron. J., 145, 101, 2013

work page 2013

[5] [4]

Bull., 67, 115, 2012

E.I.Kaisina, D.I.Makarov, I.D.Karachentsev, S.S.Kaisin, Astrophys . Bull., 67, 115, 2012

work page 2012

[6] [5]

J., 676, 184, 2008

R.B.Tully, E.J.Shaya, I.D.Karachentsev et al., Astrophys. J., 676, 184, 2008

work page 2008

[7] [6]

P. J. E.Peebles, S.D.Phelps, E.J.Shaya, R.B.Tully, Astrophys. J., 554, 104, 2001

work page 2001

[8] [7]

J., 850, 207, 2017

E.J.Shaya, R.B.Tully, Y.Hoﬀman, D.Pomar´ ede, Astrophys. J., 850, 207, 2017

work page 2017