pith. sign in

arxiv: 2606.10655 · v1 · pith:FM3G573Dnew · submitted 2026-06-09 · 🌌 astro-ph.GA

Optical-morphology-based assessment of astrometric quality in Gaia-CRF3 quasars

Qiqi Wu , Shilong Liao , Zhaoxiang Qi , Qi Xu , Ye Ding , Keyu Zhu This is my paper

Pith reviewed 2026-06-27 12:52 UTC · model grok-4.3

classification 🌌 astro-ph.GA
keywords Gaia-CRF3quasarsastrometric systematicsmorphological scoreparallax zero-pointcelestial reference frameproper motionpoint-source selection
0
0 comments X

The pith

AGNs with non-point-like morphology in Gaia-CRF3 produce a -43.7 μas parallax zero-point shift that standard corrections fail to remove.

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

The paper constructs a Gaia-independent morphological score for 1.6 million Gaia-CRF3 sources by training an XGBoost model on parameters from DESI, SDSS, SkyMapper, and the PS1 point-source catalog, then fusing the outputs into a single 0-to-1 composite indicator. It reports that sources with strongly extended morphology induce a parallax zero-point offset of approximately -43.7 microarcseconds. The same indicator reveals that subsamples with different morphological purity produce measurably different all-sky proper-motion fields. Selecting only sources with point_score above 0.95 reduces the total frame spin amplitude by 15.8 percent compared with the full catalog.

Core claim

Using a multi-survey fusion of morphological scores derived via XGBoost, the authors demonstrate that AGNs with strongly non-point-like morphology induce a parallax zero-point shift of about -43.7 μas in Gaia-CRF3, which cannot be effectively removed by the current correction model. Reference-source subsamples selected in different score ranges exhibit significantly different all-sky proper motion fields; the high-purity point-source subsample with point_score > 0.95 reduces the total frame spin amplitude by 15.8 percent relative to the full Gaia-CRF3 sample.

What carries the argument

The composite point-source score, a 0-to-1 Gaia-independent indicator of departure from ideal point-source morphology obtained by fusing XGBoost predictions trained on external survey morphological parameters.

If this is right

  • Morphology-aware selection can reduce systematic errors in the Gaia celestial reference frame.
  • The current parallax zero-point model requires additional terms that depend on source extent.
  • Proper-motion fields derived from reference sources vary with the morphological purity of the sample.
  • High-purity point-source subsamples improve the rotational stability of the reference frame by 15.8 percent.

Where Pith is reading between the lines

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

  • Future reference-frame constructions may benefit from routine morphological filtering before spin or zero-point fitting.
  • The same external-score approach could be applied to other large astrometric catalogs to diagnose similar biases.
  • If extended morphology correlates with host-galaxy properties, the effect may also appear in studies of AGN proper motions over longer baselines.

Load-bearing premise

The XGBoost model trained on morphological parameters from other surveys produces scores that accurately measure true optical extent without inheriting or masking Gaia-specific systematics.

What would settle it

Repeating the parallax zero-point measurement for the same strongly non-point-like AGN subsample using an independent astrometric catalog or high-resolution imaging would test whether the -43.7 μas offset persists.

Figures

Figures reproduced from arXiv: 2606.10655 by Keyu Zhu, Qiqi Wu, Qi Xu, Shilong Liao, Ye Ding, Zhaoxiang Qi.

Figure 1
Figure 1. Figure 1: normalised density score distributions by SDSS morpho [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Distribution of Gaia-CRF3 sources by the number of [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Distribution of the point_score of Gaia-CRF3 sources. Article number, page 5 [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: Distribution of the corrected BP/RP flux-excess fac￾tor, C ∗ , versus point_score for Gaia-CRF3 sources. Each point represents one source, and darker colours indicate higher source density. The blue curve shows the median value in each point_score bin. 15 16 17 18 19 20 21 phot_g_mean_mag / mag 0 0.2 0.4 0.6 0.8 1.0 point_score [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: presents the distribution of point_score as a function of Gaia G magnitude, together with the trend of the median value. Although the median begins to decline beyond G ∼ 19, it re￾mains above 0.9 for G < 20.85, indicating that the score re￾mains effective over this magnitude range. For G > 20.85, the median drops rapidly. This is partly because the number of Gaia sources becomes very small in this regime, … view at source ↗
Figure 7
Figure 7. Figure 7: Heat maps of the deviations of the normalised Gaia astrometric quantities from a standard normal distribution as a function [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Weighted mean parallax as a function of G magnitude for subsamples defined by different point_score. The upper and lower panels correspond to the five-parameter and six-parameter solutions, respectively. The left column shows the raw paral￾laxes, while the right column shows the results after applying the parallax zero-point correction from Lindegren et al. (2021a). The weighted mean was computed using wei… view at source ↗
Figure 9
Figure 9. Figure 9: Weighted mean proper motion systematics in bins of the [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Change in Gaia-CRF3 spin amplitude as a function of [PITH_FULL_IMAGE:figures/full_fig_p010_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: All-sky proper motion vector fields of Gaia-CRF3 sub [PITH_FULL_IMAGE:figures/full_fig_p010_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Pie chart showing the primary origin of the low point [PITH_FULL_IMAGE:figures/full_fig_p011_12.png] view at source ↗
read the original abstract

Context. Several studies have shown that host-galaxy structure or extended optical morphology in AGNs can induce spurious parallaxes and proper motions in Gaia DR3. However, it remains unclear whether source morphology also introduces systematic errors into the celestial reference frame constructed from Gaia data. Aims. We aim to provide a Gaia-independent external morphological indicator for Gaia-CRF3 sources and to use it to quantify the astrometric systematics associated with source morphology. Methods. Using morphological parameters derived from DESI, SDSS, and SkyMapper, together with the PS1-PSC point-source score as a common reference scale, we used XGBoost to infer external morphological scores for Gaia-CRF3 sources. We then developed a multi-survey fusion scheme to combine the four survey-based point-source scores into a single composite score that measures the degree to which each source departs from the morphology of an ideal point source. Results. We obtained morphological scores for 1,607,490 Gaia-CRF3 sources, corresponding to a completeness of 99.59\% with respect to the full Gaia-CRF3 catalogue. The score ranges from 0 to 1 and remains reliable for sources with $G<20.85$ mag. Based on this indicator, we find that AGNs with strongly non-point-like morphology induce a parallax zero-point shift of about $-43.7\,\mu$as, which cannot be effectively removed by the current parallax zero-point correction model. We also find that reference-source subsamples selected in different score ranges exhibit significantly different all-sky proper motion fields. For the high-purity point-source subsample with \texttt{point\_score} > 0.95, the total frame spin amplitude is reduced by 15.8\% relative to that of the full Gaia-CRF3 sample.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper develops a Gaia-independent morphological indicator for sources in the Gaia-CRF3 catalogue using an XGBoost model trained on parameters from DESI, SDSS, SkyMapper, and the PS1-PSC point-source score. A composite point-source score is constructed via multi-survey fusion, achieving 99.59% completeness. The authors report that AGNs with non-point-like morphology induce a parallax zero-point shift of approximately -43.7 μas that persists after standard corrections, and that selecting sources with point_score > 0.95 reduces the all-sky proper motion frame spin amplitude by 15.8% compared to the full sample.

Significance. If the composite morphology score is shown to be a clean, independent proxy for optical extent, the work would offer a practical external tool for vetting Gaia-CRF3 reference sources and mitigating morphology-driven systematics in the celestial reference frame. The reported 99.59% completeness and the quantitative impact on parallax zero-point and frame spin are potentially useful for future Gaia data releases and reference-frame construction.

major comments (3)
  1. [Abstract] Abstract: the reported parallax zero-point shift of -43.7 μas and the 15.8% spin-amplitude reduction are stated as point values with no accompanying uncertainties, error bars, or statistical significance tests, which is load-bearing for the claim that morphology induces uncorrectable systematics.
  2. [Methods (XGBoost training and multi-survey fusion)] Methods (XGBoost training and multi-survey fusion): no description is given of training/validation splits, cross-validation procedure, performance metrics, or feature-importance analysis for the XGBoost model that produces the point-source scores; this directly affects the reliability of the composite score used for all subsample selections and bias measurements.
  3. [Results (parallax and proper-motion analysis)] Results (parallax and proper-motion analysis): the assertion that the -43.7 μas shift cannot be removed by the current zero-point correction model, and that the spin reduction is morphology-driven, lacks explicit tests for independence from the morphology-model training itself or for selection biases correlated with magnitude, color, or redshift across the input surveys.
minor comments (2)
  1. [Abstract] Abstract: the G < 20.85 mag reliability limit is stated without indicating how this threshold was determined or whether it was validated against an external morphology catalog.
  2. The notation 'point_score' is introduced without a clear definition of its exact construction from the four survey scores in the abstract, which would aid readability.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address each major comment point-by-point below, indicating where revisions will be made to strengthen the paper.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the reported parallax zero-point shift of -43.7 μas and the 15.8% spin-amplitude reduction are stated as point values with no accompanying uncertainties, error bars, or statistical significance tests, which is load-bearing for the claim that morphology induces uncorrectable systematics.

    Authors: We agree that uncertainties and significance tests are important for these key results. In the revised manuscript we will report bootstrap-derived 1σ uncertainties on both the -43.7 μas shift and the 15.8% spin reduction, together with p-values from two-sample tests comparing the relevant subsamples. These additions will appear in the abstract and the corresponding results section. revision: yes

  2. Referee: [Methods (XGBoost training and multi-survey fusion)] Methods (XGBoost training and multi-survey fusion): no description is given of training/validation splits, cross-validation procedure, performance metrics, or feature-importance analysis for the XGBoost model that produces the point-source scores; this directly affects the reliability of the composite score used for all subsample selections and bias measurements.

    Authors: The current Methods section provides only a high-level description of the XGBoost model. We will expand it to include the 80/20 training/validation split, 5-fold cross-validation details, performance metrics (AUC-ROC, precision-recall, accuracy), and SHAP-based feature-importance rankings. This will directly address the reliability of the composite point-source score. revision: yes

  3. Referee: [Results (parallax and proper-motion analysis)] Results (parallax and proper-motion analysis): the assertion that the -43.7 μas shift cannot be removed by the current zero-point correction model, and that the spin reduction is morphology-driven, lacks explicit tests for independence from the morphology-model training itself or for selection biases correlated with magnitude, color, or redshift across the input surveys.

    Authors: We will add explicit checks in the revised Results section: (i) magnitude-, color-, and redshift-matched subsample comparisons to demonstrate that the observed differences are not driven by selection biases, and (ii) verification that the morphology score shows no significant correlation with Gaia astrometric parameters outside the external-survey training data. These tests will support the claim that the shift persists after the standard correction and that the spin reduction is morphology-driven. revision: partial

Circularity Check

0 steps flagged

No circularity: morphology scores derived from independent external surveys and applied to Gaia data

full rationale

The paper trains an XGBoost model on morphological parameters and point-source scores from DESI, SDSS, SkyMapper and PS1-PSC (all external to Gaia), then applies the resulting composite point_score to matched Gaia-CRF3 sources. The reported parallax offset (-43.7 μas) and spin reduction (15.8%) are obtained by splitting the Gaia sample on this external score and measuring differences in Gaia astrometry; neither quantity is defined by or forced to equal any fitted parameter or equation inside the paper. No self-citations, self-definitional loops, or fitted-input-as-prediction patterns appear in the derivation chain. The central results remain independent of the paper's own inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claim rests on the assumption that external survey morphology is a faithful proxy for the optical structure that affects Gaia astrometry, and on the validity of the machine-learning mapping; no explicit free parameters are listed beyond model training choices.

axioms (1)
  • domain assumption The PS1-PSC point-source score provides a reliable common reference scale across the four surveys for training the morphology model.
    Invoked to create a unified training target for XGBoost.
invented entities (1)
  • composite point-source score no independent evidence
    purpose: Single scalar measuring departure from ideal point-source morphology
    Constructed by fusing four survey-based scores; no independent falsifiable prediction supplied.

pith-pipeline@v0.9.1-grok · 5877 in / 1222 out tokens · 26084 ms · 2026-06-27T12:52:12.040533+00:00 · methodology

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

35 extracted references · 1 linked inside Pith

  1. [1]

    S., Gordon, D., et al

    Charlot, P., Jacobs, C. S., Gordon, D., et al. 2020, A&A, 644, A159

    2020

  2. [2]

    & Guestrin, C

    Chen, T. & Guestrin, C. 2016, arXiv e-prints, arXiv:1603.02754

    Pith/arXiv arXiv 2016

  3. [3]

    M., Gerke, B

    Comerford, J. M., Gerke, B. F., Stern, D., et al. 2012, ApJ, 753, 42

    2012

  4. [4]

    J., Lang, D., et al

    Dey, A., Schlegel, D. J., Lang, D., et al. 2019, AJ, 157, 168

    2019

  5. [5]

    2024, A&A, 691, A81

    Ding, Y ., Liao, S., Wu, Q., Qi, Z., & Tang, Z. 2024, A&A, 691, A81

    2024

  6. [6]

    2023, A&A, 674, A11

    Ducourant, C., Krone-Martins, A., Galluccio, L., et al. 2023, A&A, 674, A11

    2023

  7. [7]

    2018, A&A, 618, A56

    Ducourant, C., Wertz, O., Krone-Martins, A., et al. 2018, A&A, 618, A56

    2018

  8. [8]

    2016, A&A, 595, A3

    Fabricius, C., Bastian, U., Portell, J., et al. 2016, A&A, 595, A3

    2016

  9. [9]

    2021, A&A, 649, A5

    Fabricius, C., Luri, X., Arenou, F., et al. 2021, A&A, 649, A5

    2021

  10. [10]

    Fey, A. L. & Charlot, P. 1997, ApJS, 111, 95

    1997

  11. [11]

    L., Gordon, D., Jacobs, C

    Fey, A. L., Gordon, D., Jacobs, C. S., et al. 2015, AJ, 150, 58

    2015

  12. [12]

    J., et al

    Fu, Y ., Wu, X.-B., Bouwens, R. J., et al. 2025, ApJS, 279, 54 Gaia Collaboration, Bailer-Jones, C. A. L., Teyssier, D., et al. 2023, A&A, 674, A41 Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al. 2021, A&A, 649, A1 Gaia Collaboration, Klioner, S. A., Lindegren, L., et al. 2022, A&A, 667, A148 Gaia Collaboration, Krone-Martins, A., Ducourant, C....

    2025

  13. [13]

    2020, ApJ, 888, 73

    Hwang, H.-C., Shen, Y ., Zakamska, N., & Liu, X. 2020, ApJ, 888, 73

    2020

  14. [14]

    2023, MNRAS, 524, 1909

    Ji, X., Zheng, Z.-Y ., Wu, Q., et al. 2023, MNRAS, 524, 1909

    2023

  15. [15]

    E., et al

    Kaiser, N., Aussel, H., Burke, B. E., et al. 2002, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, V ol. 4836, Survey and Other Telescope Technologies and Discoveries, ed. J. A. Tyson & S. Wolff, 154–164

    2002

  16. [16]

    A., Auger, M

    Lemon, C. A., Auger, M. W., & McMahon, R. G. 2019, MNRAS, 483, 4242

    2019

  17. [17]

    2025, arXiv e-prints, arXiv:2511.02204

    Liao, S., Wu, Q., Ding, Y ., Xu, Q., & Qi, Z. 2025, arXiv e-prints, arXiv:2511.02204

    arXiv 2025

  18. [18]

    2021, PASP, 133, 094501

    Liao, S., Wu, Q., Qi, Z., et al. 2021, PASP, 133, 094501

    2021

  19. [19]

    F., Eubanks, T

    Ma, C., Arias, E. F., Eubanks, T. M., et al. 1998, AJ, 116, 516

    1998

  20. [20]

    2025, Nature Astronomy, 9, 1396

    Makarov, V . 2025, Nature Astronomy, 9, 1396

    2025

  21. [21]

    Makarov, V . V . & Secrest, N. J. 2022, ApJ, 933, 28

    2022

  22. [22]

    Makarov, V . V . & Secrest, N. J. 2023, ApJS, 264, 4

    2023

  23. [23]

    2022, Nature Astronomy, 6, 1185 Massey Jr, F

    Mannucci, F., Pancino, E., Belfiore, F., et al. 2022, Nature Astronomy, 6, 1185 Massey Jr, F. J. 1951, Journal of the American statistical Association, 46, 68

    2022

  24. [24]

    & Klioner, S

    Mignard, F. & Klioner, S. 2012, A&A, 547, A59

    2012

  25. [25]

    Miller, A. A. & Hall, X. J. 2021, PASP, 133, 054502

    2021

  26. [26]

    & Kovalev, Y

    Petrov, L. & Kovalev, Y . Y . 2017, MNRAS, 471, 3775

    2017

  27. [27]

    W., Weaver, K

    Pfeifle, R. W., Weaver, K. A., Secrest, N. J., Rothberg, B., & Patton, D. R. 2025, ApJS, 281, 25 Popovi´c, L. ˇC., Jovanovi´c, P., Stalevski, M., et al. 2012, A&A, 538, A107

    2025

  28. [28]

    E., Berghea, C

    Rusu, C. E., Berghea, C. T., Fassnacht, C. D., et al. 2019, MNRAS, 486, 4987

    2019

  29. [29]

    & Miller, A

    Tachibana, Y . & Miller, A. A. 2018, PASP, 130, 128001

    2018

  30. [30]

    Taylor, M. B. 2005, in Astronomical Society of the Pacific Conference Se- ries, V ol. 347, Astronomical Data Analysis Software and Systems XIV , ed. P. Shopbell, M. Britton, & R. Ebert, 29

    2005

  31. [31]

    A., Luvaul, L

    Wolf, C., Onken, C. A., Luvaul, L. C., et al. 2018, PASA, 35, e010

    2018

  32. [32]

    2023, Research in Astronomy and Astrophysics, 23, 025006

    Wu, Q., Liao, S., Qi, Z., et al. 2023, Research in Astronomy and Astrophysics, 23, 025006

    2023

  33. [33]

    2024, A&A, 692, A154

    Wu, Q., Scialpi, M., Liao, S., Mannucci, F., & Qi, Z. 2024, A&A, 692, A154

    2024

  34. [34]

    2022, Frontiers in Astronomy and Space Sciences, 9, 822768

    Wu, Q.-Q., Liao, S.-L., Ji, X., et al. 2022, Frontiers in Astronomy and Space Sciences, 9, 822768

    2022

  35. [35]

    G., Adelman, J., Anderson, Jr., J

    York, D. G., Adelman, J., Anderson, Jr., J. E., et al. 2000, AJ, 120, 1579 Article number, page 13 A&A proofs:manuscript no. quality_qso Appendix A: normalised astrometric distributions in Morphological Score and magnitude bins Fig. A.1 shows the distributions of the normalised Gaia astro- metric quantitiesϖ/σ ϖ,µ α∗/σµα∗, andµ δ/σµδ for the Gaia- CRF3 so...

    2000