REVIEW
CosmoVerse Data Challenge: How Tip of the Red Giant Branch Analysis Choices Shape the Inferred Hubble Constant
Astrophysics of Galaxies · Submitted 3 Nov 2026 · 127 references
Pith's verdict: Solid, well-executed community data challenge that quantifies spread in TRGB-based H0 from analysis choices; limited novelty but high reproducibility and clear value for the subfield.
Why Pith says this
Highlight any text to comment on it.
Referee report
The paper
The paper claims
TRGB analysis choices alone can shift the inferred Hubble constant by tens of km/s/Mpc, as shown by a community data challenge where 26 submissions produced H0 values ranging from 67.91 to 99.32 km/s/Mpc.
The reading
What carries the argument
The TRGB measurement pipeline combines GLOESS smoothing of the luminosity function, Sobel edge detection, and bootstrap resampling to identify the tip magnitude. Participants independently set the smoothing bandwidth, color-selection window, spatial-clipping radius, and edge-detection weighting for each galaxy and for the anchor NGC 4258. The submitted distances are then fed into a shared SN Ia notebook that uses the Pantheon+ covariance matrix and a generalized least-squares solver to produce H0, holding fixed the SN data and fitting procedure. This design isolates the effect of TRGB analysis choices on the final H0 value.
Extended reading notes
Core claim
Under a fixed SN Ia framework (Pantheon+ dataset and fitting code), variation in TRGB methodology alone produces H0 values that scatter by 7.38 km/s/Mpc (standard deviation) among submissions, with a tail extending to 99 km/s/Mpc. Two factors dominate: (1) Inclusion of four galaxies (NGC 1309, NGC 3021, NGC 3370, NGC 5584) with low tip contrast and systemically smaller distance moduli—when any of them enters the fit, H0 rises by a median of 6.5 km/s/Mpc; (2) The anchor tip magnitude of NGC 4258, which differs by 0.079 mag (standard deviation) between submissions and shifts all galaxy distances coherntly. Changing smoothing bandwidth, color cuts, and spatial clipping also affects individual distances, but the largest portion of the variance traces back to galaxy selection and anchor measurement. The paper shows that reasonable methodological choices within a single framework can produce substantially different H0 values, and it makes all data and code public so that the differences can be inspected and traced.
Load-bearing premise
The central claim that TRGB analysis choices produce the observed spread in H0 depends on the assumption that the fixed SN Ia notebook and Pantheon+ dataset correctly propagate TRGB distances into H0, and that the participants' choices span the range of reasonable TRGB methodology. The paper does not establish that any particular choice is more correct; it only characterizes the spread. The weakest load-bearing premise is that the four galaxies with systematically low tip contrast (NGC 1309, NGC 3021, NGC 3370, NGC 5584) indeed have biased TRGB distances, which is inferred from comparison with Cepheid distances (not from an independent TRGB measurement) and from the fact that many participants obtained low tip contrast in those galaxies.
Editorial extensions
If this is right
- If calibrator galaxy selection is as influential as found here, published TRGB-based H0 values that differ by 5–10 km/s/Mpc may be partly reconciled by examining which hosts are included.
- The anchor galaxy NGC 4258 requires a standardized, multi-field tip measurement because field-to-field scatter of ~0.08 mag propagates into a ~3 km/s/Mpc spread in H0.
- Future TRGB analyses should report a systematic uncertainty term that accounts for the choice of calibrator subset and for the sensitivity of the tip detection to smoothing and color cuts.
- Community data challenges of this type provide a reproducible framework for quantifying methodological dispersion and can be extended to other rungs of the distance ladder or to blinded rounds.
Reading between the lines
- The paper’s results imply that part of the 6–7 km/s/Mpc gap between local and CMB-inferred H0 may be an artifact of particular calibrator choices: if the four low-contrast galaxies are excluded, the median of the submissions drops from 77.1 to 76.0 km/s/Mpc, moving closer to the Cepheid-based SH0ES value of 73.0 km/s/Mpc.
- The wide spread also highlights that the ‘tip of the red giant branch’ method, while physically well-motivated, is highly operator-dependent in practice; a community-endorsed standardization protocol (e.g., mandatory tip-contrast cuts and a fixed anchor measurement) would reduce this variance.
- Because the same fixed SN Ia notebook was used for all submissions, any hidden systematic in the Pantheon+ analysis (e.g., host-mass step, extinction law) would shift all submissions together and cannot explain the observed spread, strengthening the paper’s attribution to TRGB choices.
Original abstract
Tip of the red giant branch (TRGB)-based determinations of the Hubble constant ($H_0$), the current expansion rate of the Universe, depend on choices made in selecting and analyzing stellar photometry and in assembling the sample of calibrator galaxies hosting Type Ia supernovae (SNe Ia). The CosmoVerse Data Challenge invites participants to analyze public photometric catalogs within a shared distance-ladder framework, opening each step of the measurement to the community and testing how these choices propagate into $H_0$. Participants independently selected photometric catalogs, galaxies, and TRGB measurement settings, while using a common SN Ia compilation and fitting procedure to convert distances into $H_0$. Supported by public notebooks and a training school, the challenge received 26 submissions from 22 participants. All submissions used the same analysis notebook, many participants were trainees at the school, and the true galaxy distances are unknown; the results therefore characterize the sensitivity of the submissions to analysis and sample-selection choices and are not a combined determination of $H_0$ or an estimate of the systematic uncertainty in published TRGB measurements. The inferred $H_0$ values have a median of $77.10\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ and a standard deviation of $7.38\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ $(5.11\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ for the scaled median absolute deviation), $\sim 3.4$ $(\sim 2.4)$ times the typical quoted uncertainty of a submission. Differences in the measured tip of the anchor galaxy NGC 4258 add a coherent shift in the distance scale. The challenge shows how different choices within a common framework can produce substantially different values of $H_0$, and provides an open basis for tracing these differences and for making datasets and tools more transparent and accessible.
The compiled record
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
The abstract, sentence by sentence
Tip of the red giant branch (TRGB)-based determinations of the Hubble constant ($H_0$), the current expansion rate of the Universe, depend on choices made in selecting and analyzing stellar photometry and in assembling the sample of calibrator galaxies hosting Type Ia supernovae (SNe Ia). The CosmoVerse Data Challenge invites participants to analyze public photometric catalogs within a shared distance-ladder framework, opening each step of the measurement to the community and testing how these choices propagate into $H_0$. Participants independently selected photometric catalogs, galaxies, and TRGB measurement settings, while using a common SN Ia compilation and fitting procedure to convert distances into $H_0$. Supported by public notebooks and a training school, the challenge received 26 submissions from 22 participants. All submissions used the same analysis notebook, many participants were trainees at the school, and the true galaxy distances are unknown; the results therefore characterize the sensitivity of the submissions to analysis and sample-selection choices and are not a combined determination of $H_0$ or an estimate of the systematic uncertainty in published TRGB measurements. The inferred $H_0$ values have a median of $77.10\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ and a standard deviation of $7.38\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ $(5.11\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ for the scaled median absolute deviation), $\sim 3.4$ $(\sim 2.4)$ times the typical quoted uncertainty of a submission. Differences in the measured tip of the anchor galaxy NGC 4258 add a coherent shift in the distance scale. The challenge shows how different choices within a common framework can produce substantially different values of $H_0$, and provides an open basis for tracing these differences and for making datasets and tools more transparent and accessible.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Tip of the red giant branch (TRGB)-based determinations of the Hubble constant ($H_0$), the current expansion rate of the Universe, depend on choices made in selecting and analyzing stellar photometry and in assembling the sample of calibrator galaxies hosting Type Ia supernovae (SNe Ia). | background 1.00 | 0.07 | 0.05 | asserted 1.00 |
| s1 | The CosmoVerse Data Challenge invites participants to analyze public photometric catalogs within a shared distance-ladder framework, opening each step of the measurement to the community and testing how these choices propagate into $H_0$. | contribution 0.84 | 0.48 | 0.05 | asserted 0.93 |
| s2 | Participants independently selected photometric catalogs, galaxies, and TRGB measurement settings, while using a common SN Ia compilation and fitting procedure to convert distances into $H_0$. | method 0.98 | 0.43 | 0.05 | asserted 1.00 |
| s3 | Supported by public notebooks and a training school, the challenge received 26 submissions from 22 participants. | method 0.72 | 0.77 | 0.02 | asserted 1.00 |
| s4 | All submissions used the same analysis notebook, many participants were trainees at the school, and the true galaxy distances are unknown; the results therefore characterize the sensitivity of the submissions to analysis and sample-selection choices and are not a combined determination of $H_0$ or an estimate of the systematic uncertainty in published TRGB measurements. | limitation 0.96 | 0.63 | 0.19 | asserted 0.99 |
| s5 | The inferred $H_0$ values have a median of $77.10\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ and a standard deviation of $7.38\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ $(5.11\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ for the scaled median absolute deviation), $\sim 3.4$ $(\sim 2.4)$ times the typical quoted uncertainty of a submission. | result 1.00 | 0.87 | 0.19 | asserted 1.00 |
| s6 | Differences in the measured tip of the anchor galaxy NGC 4258 add a coherent shift in the distance scale. | result 0.64 | 0.50 | 0.06 | asserted 0.99 |
| s7 | The challenge shows how different choices within a common framework can produce substantially different values of $H_0$, and provides an open basis for tracing these differences and for making datasets and tools more transparent and accessible. | interpretation 0.97 | 0.79 | 0.79 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
Figures
Figures reproduced from arXiv:2611.05468 by the authors.
References
Discussion
Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.