Pith. sign in

REVIEW 4 major objections 5 minor 41 references

An automated probabilistic pipeline makes prediscovery detections of near-Earth asteroids in archival survey images, extending orbits by years.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-04 10:54 UTC pith:SZKXMZZR

load-bearing objection Real prediscovery detections, not a toy; the significance calibration at large search volumes and missing code/data are the things to fix, not the core result. the 4 major comments →

arxiv 2510.07588 v2 pith:SZKXMZZR submitted 2025-10-08 astro-ph.EP astro-ph.IM

An Automated Probabilistic Asteroid Prediscovery Pipeline

Sage Li , Alex Geringer-Sameth , Nathan Golovich This is my paper
classification astro-ph.EP astro-ph.IM
keywords prediscovery detectionnear-Earth asteroidsorbital covariance propagationZwicky Transient Facilitylikelihood ratio linkingarchival survey imagesplanetary defenseorbit determination
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper claims that prediscovery detections of near-Earth asteroids—serendipitous images taken before official discovery—can be found automatically and probabilistically in archival survey data. The authors propagate the full orbital covariance from a fit to post-discovery astrometry into archival ZTF images, search low-threshold source catalogs within the resulting sky regions, and use a likelihood-ratio linking step to separate true detections from the large number of false candidates. They report recovering the potentially hazardous asteroid 2021 DG1 in ZTF images from 2.5 years before its discovery, shrinking its 2035 sky-plane uncertainty from many degrees to arcseconds, and recovering 2025 FU24 nearly 7 years before its first known observation. If the method holds up, it turns already-collected survey images into a resource for immediately refining the orbits of newly discovered near-Earth objects, which matters for impact-risk assessment. The pipeline is survey-agnostic and designed for the coming flood of discoveries from Rubin, NEO Surveyor, and NEOMIR.

Core claim

The central claim is that a pipeline combining orbit refitting, full-covariance propagation, and a probabilistic linking statistic can make genuine prediscovery detections of near-Earth asteroids in archival survey images. For each NEA, the authors refit post-discovery astrometric observations to obtain a best-fit orbit and six-parameter covariance, then propagate 20,000 samples from the 99.99% confidence surface to every survey epoch, constructing a convex-hull search region from the sample locations. Every low-threshold source detected inside that region is treated as a candidate; each candidate is appended to the post-discovery data, refit, and the resulting trial orbit propagated to all

What carries the argument

The load-bearing object is the orbital covariance: a six-parameter Keplerian covariance from a least-squares fit to post-discovery astrometry, propagated into each survey image. The propagation is done two ways: a Monte Carlo draw of 20,000 orbits from the 99.99% chi-squared surface to define convex-hull search regions, and a fast first-order covariance mapping (Appendix A) that uses only 12 finite-difference integrations to estimate per-image uncertainty for the linking step. The linking statistic is a likelihood ratio comparing, for each candidate source, the observed distribution of Mahalanobis distances to the closest source in every other image under the alternative (the candidate is th

Load-bearing premise

The 99.99%-level covariance ellipsoid propagated from the post-discovery fit is trusted to contain the true pre-discovery position in every archival image, and any mis-sized per-image covariance would cause the asteroid to be missed or a false positive to rank above it.

What would settle it

For a set of near-Earth asteroids with independently known precovery detections in archival images, truncate the post-discovery arc to just the discovery observations, run the pipeline, and check whether the true precovery position falls inside the constructed search region in every image; if it falls outside in any image, the covariance-coverage assumption fails.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • For 2021 DG1, extending the arc to 2.5 years reduces the 2035 sky-plane uncertainty from many degrees to arcseconds, making recovery straightforward.
  • For 2025 FU24, the arc grows by a factor of 78, from 31 days to nearly 7 years.
  • The pipeline can be applied immediately to any new NEA discovery, searching archival images within the constraints of the propagated covariance.
  • About 500 of the roughly 3,000 NEAs identified as precovery candidates could have their arcs doubled using data already on disk.
  • The method is survey-agnostic and scales to future facilities like Rubin, NEO Surveyor, and NEOMIR, where the discovery rate will sharply increase.

Where Pith is reading between the lines

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

  • The pipeline's effectiveness hinges on the accuracy of the initial covariance; if the post-discovery orbit is systematically biased, no amount of linking will help. A natural extension is to run the pipeline on objects whose precovery is already known to measure true-positive rate as a function of uncertainty parameter U.
  • The likelihood-ratio framework could be extended to incorporate non-detections as Bayesian constraints, using null images to rule out portions of the orbit space—the approach hinted at for 2024 YR4.
  • The constant source-density assumption will break down in crowded fields near bright artifacts; modeling position-dependent density could reduce false positives without sacrificing sensitivity.
  • The 500-pixel buffer that truncates banana-shaped uncertainty regions may cause the pipeline to miss the oldest, most valuable precoveries; a curved-region search that handles the full covariance without truncation could push the method to longer arcs.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper describes an automated pipeline for finding prediscovery detections of near-Earth asteroids in archival survey images. Starting from MPC astrometry, the authors refit a six-parameter Keplerian orbit and covariance, propagate 20,000 samples to ZTF epochs, define convex-hull search regions, build low-threshold source catalogs, and for every candidate source refit an orbit and compute a likelihood ratio linking that source to sources in other images. They demonstrate the method on two NEAs: 2021 DG1, with 19 catalog sources within 1 arcsec of the predicted position and an arc extension of 2.5 yr, and 2025 FU24, recovered nearly 7 yr before discovery with 18 inlier sources. They also present two null tests on faint objects and argue that the pipeline identifies ~2,676 NEAs with prediscovery potential, including ~490 whose arcs could be doubled.

Significance. The claimed results are scientifically valuable: genuine prediscovery detections derived from backward propagation of post-discovery astrometry, rather than from fitted quantities, would substantially improve orbit determination for newly discovered NEAs and inform follow-up for objects such as 2024 YR4. The paper's strengths include a clearly stated probabilistic framework, explicit hypotheses for the likelihood ratio, self-consistency checks on the significant candidates, and the fact that the search regions are genuine predictions from MPC data independent of the archival detections. The two positive detections pass an internal consistency check (multiple images aligned along one orbit). However, the central significance statistic relies on an approximation whose validity is not demonstrated in the regime of the headline detection, and the paper provides no code, data products, or empirical false-positive-rate measurement across the full candidate sample. These are load-bearing for the claim that the >10-sigma detections are real and that the method can be trusted at scale.

major comments (4)
  1. [§3.5 and Appendix A] The calibration of the detection significance rests on Eq. B9, which uses Mahalanobis distances from Eq. 7 and the per-image covariance S^{mm'} defined in Eq. 6. This covariance is estimated once from an exemplar trial orbit via a first-order Jacobian (Appendix A) and then reused for all trial sources from the same image. That assumption is unlikely to hold in the long-arc, large-search-volume regime: the paper itself notes banana-shaped uncertainty regions and truncation by a 500-pixel buffer in §3.2. For 2025 FU24, the search spans 5,048 images and 184,452 sources; a mis-sized or misshapen covariance directly biases the likelihood ratios and can inflate the standardized significance. I request a validation: draw Monte Carlo samples from the full refit covariance for a subset of trial orbits, propagate them, and compare the empirical spread with the single Gaussian approximation, or oth
  2. [§4.4 and §3.5] The null tests are far smaller than the positive detection searches. 2022 DB4 uses 833 images/4,943 sources and 2022 ED1 uses 812 images/694 sources, whereas 2025 FU24 uses 5,048 images/184,452 sources. Finding no >10-sigma outliers in the small null tests does not establish the false-positive rate over ~2,676 objects in §4.1. The paper should report the empirical distribution of the maximum standardized significance under the null hypothesis across a large set of faint objects, or an analytic estimate of the expected number of false positives above the adopted threshold. Without this, the statement that 'a simple threshold like significance >10 easily separates prediscoveries' is not supported at the scale claimed.
  3. [Reproducibility / data availability] The manuscript does not provide the pipeline code, the generated source catalogs, the trial-orbit lists, or the exact data versions used for the two positive detections. Given the number of hyperparameters (Huber c, alpha, source threshold, buffer size, magnitude softening) and the complexity of the linking procedure, independent verification is currently impossible. At a minimum, the authors should release code and the per-image covariance estimates, and list the archival ZTF product identifiers for the 19 and 18 matched images, so that the claimed detections can be checked.
  4. [§3.1 and hyperparameters] Several choices are described as manual or tuned without sensitivity analysis: the Huber scale c=1, the prior alpha=0.5, the 3-sigma detection threshold softened by one magnitude, and the 500-pixel buffer for outlier removal. These choices affect the search regions and the likelihood ratios. I ask for a brief sensitivity study, e.g., varying alpha between 0.1 and 0.9 and the buffer between 100 and 1000 pixels, to show that the two positive detections are not artifacts of particular hyperparameter values.
minor comments (5)
  1. [Figure 5 caption] The bottom-right panel is labeled '2022 ED4' in the caption but the text and Figure 4 refer to 2022 ED1. Please correct.
  2. [Figure 6 and §4.2] The text says samples were propagated 'to 2025-07-31', while the figure caption says the epoch is 2035-07-31. This is likely a typo, but it should be clarified because the numerical sky-plane error depends on the epoch.
  3. [§3.4, Eq. (3)] The sum over j=1 to 2 is notationally unclear because the argument of chi^2 does not contain an index j. Please expand the two terms explicitly or define the RA/Dec components.
  4. [§4.2 and §5] The paper shifts between 'prediscovery' and 'predetection' (see abstract/discussion); use one term consistently. Also, 'Feburary' on page 7 is misspelled.
  5. [Appendix B.1] The derivation of the null distribution assumes independent images and a homogeneous Poisson process. This is stated as an approximation, but the clumping of sources around artifacts is acknowledged. A sentence quantifying the impact of source clustering on the variance in Eq. B9 would be helpful.

Circularity Check

0 steps flagged

No significant circularity: precovery search regions are propagated from independent MPC astrometry; archival detections are not used to define the test.

full rationale

The paper's derivation chain is self-contained. Search regions are constructed by fitting MPC post-discovery astrometry (Eqs. 1-2), propagating the full 6-parameter covariance to survey epochs, and sampling the 99.99% confidence ellipsoid to build convex hulls (Sec. 3.2). The candidate sources in those regions are detected independently in archival ZTF images (Sec. 3.3). The linking statistic (Eqs. 5-9, B5-B9) compares Mahalanobis distances against a null point-process model whose parameters are the image source density and the propagated covariance; no fitted quantity from the candidate detections is used to define the null or the significance. The per-image covariance approximation in Appendix A is a modeling assumption, not a circular reduction: it is computed from the post-discovery fit's Jacobian and applied uniformly to all trial sources, so it is not calibrated on the sources it later evaluates. The claimed prediscoveries of 2021 DG1 and 2025 FU24 are validated by self-consistent cross-image linking (Sec. 3.6) and by null tests on objects too faint to be detected (Sec. 4.4). The self-citations that appear (e.g., Geringer-Sameth et al. 2015 for Fourier convolution methods; Golovich et al. 2025 and Geringer-Sameth et al. 2025 for stacking searches) are methodological pointers in non-load-bearing contexts, not premises that define the target result. No equation reduces to its own output, and no fitted parameter is renamed as a prediction. Thus the central claims have independent empirical content and the paper does not exhibit circularity.

Axiom & Free-Parameter Ledger

6 free parameters · 5 axioms · 0 invented entities

The paper introduces no new physical entities, particles, or forces; it is a purely computational pipeline. The free parameters listed are hand-tuned or assumed and, except for the detection thresholds, do not directly determine the strongest claim (that the pipeline finds true prediscoveries), but they affect the search regions and significance calibration.

free parameters (6)
  • Huber factor c = 1
    §3.1: chosen through manual testing to balance convergence and robustness; affects the post-discovery orbital fit and covariance.
  • MPC astrometric error sigma_i = 2 arcsec (default)
    §3.1: assigned to observations without reported uncertainty; conservative but arbitrary.
  • Detection threshold softening = 3σ + 1 mag
    §3.2: images rejected if predicted flux below 3σ point-source threshold softened by 1 magnitude; hand-set.
  • Source detection threshold = 4σ
    §3.3: DAOFIND threshold; authors note reducing it is likely beneficial.
  • Buffer for hull outlier removal = 500 pixels
    §3.2: samples outside 500-pixel buffer removed to avoid banana-shaped hulls.
  • alpha prior = 0.5
    §3.5/Appendix B: prior probability the asteroid is detected in a given image; 'we found good results'.
axioms (5)
  • domain assumption ASSIST/IAS15 with DE441 provides sufficiently accurate ephemerides for backward propagation over years.
    §3.1: used for all orbit propagation; no independent validation in this paper.
  • domain assumption MPC astrometric errors, after Huber weighting, are approximately Gaussian so that the Fisher-information covariance and the 99.99% ellipsoid bracket the true orbit.
    §3.1, §3.2: if covariance is underestimated, search regions may miss the true position.
  • ad hoc to paper Background sources in ZTF difference images are well approximated by a homogeneous Poisson process.
    Appendix B: the null distribution of the likelihood ratio assumes constant density ρ; the paper notes this is an 'oversimplification' for clumped artifacts.
  • domain assumption The first-order (linearized) covariance propagation in Appendix A is sufficient to represent the uncertainty of trial orbits in destination images.
    Appendix A: uses 12 integrations instead of Monte Carlo; assumes linearity over the relevant phase space.
  • domain assumption The H-G model and MPC-reported H/G values predict the asteroid's brightness well enough for image selection.
    §3.2: light-curve model used to reject images; rotational variation is softened by 1 mag.

pith-pipeline@v1.3.0-alltime-deepseek · 15722 in / 13222 out tokens · 114234 ms · 2026-08-04T10:54:38.927246+00:00 · methodology

0 comments
read the original abstract

We present an automated and probabilistic method to make prediscovery detections of near-Earth asteroids (NEAs) in archival survey images, with the goal of reducing orbital uncertainty immediately after discovery. We refit the Minor Planet Center's astrometry and propagate the full six-parameter covariance to survey epochs to define search regions. We build low-threshold source catalogs for viable images and evaluate every detected source in a search region as a candidate prediscovery. We eliminate false positives by refitting a new orbit to each candidate and probabilistically linking detections across images using a likelihood ratio. Applied to Zwicky Transient Facility (ZTF) imaging, we identify approximately 3000 recently discovered NEAs with prediscovery potential, including a doubling of the observational arc for about 500. We use archival ZTF imaging to make prediscovery detections of the potentially hazardous asteroid 2021 DG1, extending its arc by 2.5 yr and reducing future apparition sky plane uncertainty from many degrees to arcseconds. We also recover 2025 FU24 nearly 7 yr before its first known observation, when its sky plane uncertainty covers hundreds of square degrees across thousands of ZTF images. The method is survey agnostic and scalable, enabling rapid orbit refinement for new discoveries from Rubin, NEO Surveyor, and NEOMIR.

Figures

Figures reproduced from arXiv: 2510.07588 by Alex Geringer-Sameth, Nathan Golovich, Sage Li.

Figure 1
Figure 1. Figure 1: A visualization of the prediscovery algorithm. The orbital uncertainty from post-discovery observations is propagated back in time to the survey to identify images and search regions within those images (black ellipses). Sources detected in the search regions are candidate prediscoveries. Each candidate source corresponds to a trial orbit, which is propagated to the other images to look for coincident dete… view at source ↗
Figure 2
Figure 2. Figure 2: Prediscovery of the near-Earth asteroid 2021 DG1 in a ZTF image from August 2018, 2.5 years before its discovery date. Black points are locations of sample orbits consistent with the post-discovery observations. They are used to construct a convex hull (gray region) which forms the search region for this image. Image sources detected above 4σ significance are shown as green circles (filled if they are in t… view at source ↗
Figure 3
Figure 3. Figure 3: Potential arc extension ratios of 2,676 NEAs queried from JPL SBDB with at least 5 candidate images in ZTF. Arc extension ratio is defined as the potential in￾crease in arc divided by the current arc. ability of archival wide-field survey images years in ad￾vance offers the potential to immediately improve initial orbit determination for detections made with the new surveys coming online. 4.2. Prediscovery… view at source ↗
Figure 4
Figure 4. Figure 4: Predicted light curves of selected NEAs based on their post-discovery orbit fits. Gray points show all ZTF images that should contain the asteroid. Candidate intersec￾tions (blue) are ZTF images in which the asteroid is predicted to be brighter than the image’s 3σ detection threshold. The presence of candidate intersections before the discovery date allows for significant arc extension. The points circled … view at source ↗
Figure 6
Figure 6. Figure 6: Upper panel: Sky plane error for 2021 DG1 on 2035-07-31, which corresponds to the first time it is brighter than V = 23. Fits of its orbit with (red) and without (blue) ZTF prediscoveries were propagated with a Keplerian prop￾agator. Lower panel: zoom in on the sky-plane error for the combined fit. that intersect each other (see Sec. 3.6). For the high￾est log-likelihood ratio orbit, there are 19 images wi… view at source ↗
Figure 5
Figure 5. Figure 5: Prediscovery significance of candidate sources obtained from our pipeline for selected NEOs. Each panel shows a histogram of the standardized log-likelihood ratio (in units of standard deviations above the null hypothesis expec￾tation) for orbits corresponding to trial sources detected in search regions. There is clear evidence of prediscovery for 2021 DG1 and 2025 FU24, where orbits successfully predict t… view at source ↗

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

41 extracted references · 11 canonical work pages

  1. [1]

    P., Tollerud, E

    Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, A33, doi: 10.1051/0004-6361/201322068 Astropy Collaboration, Price-Whelan, A. M., Sip˝ ocz, B. M., et al. 2018, AJ, 156, 123, doi: 10.3847/1538-3881/aabc4f Astropy Collaboration, Price-Whelan, A. M., Lim, P. L., et al. 2022, ApJ, 935, 167, doi: 10.3847/1538-4357/ac7c74

    doi:10.1051/0004-6361/201322068 2013

  2. [2]

    B., Dobkin, D

    Barber, C. B., Dobkin, D. P., & Huhdanpaa, H. 1996, ACM Trans. Math. Softw., 22, 469–483, doi: 10.1145/235815.235821

    arXiv 1996

  3. [3]

    C., Kulkarni, S

    Bellm, E. C., Kulkarni, S. R., Graham, M. J., et al. 2019, PASP, 131, 018002, doi: 10.1088/1538-3873/aaecbe

    doi:10.1088/1538-3873/aaecbe 2019

  4. [4]

    2000, Planetary and Space Science, 48, 939, doi: 10.1016/S0032-0633(00)00060-X

    Boattini, A. 2000, Planetary and Space Science, 48, 939, doi: 10.1016/S0032-0633(00)00060-X

    work page doi:10.1016/s0032-0633(00)00060-x 2000

  5. [5]

    2001, A&A, 375, 293, doi: 10.1051/0004-6361:20010825

    Boattini, A., D’Abramo, G., Forti, G., & Gal, R. 2001, A&A, 375, 293, doi: 10.1051/0004-6361:20010825

    work page doi:10.1051/0004-6361:20010825 2001

  6. [6]

    1989, in Asteroids II, ed

    Bowell, E., Hapke, B., Domingue, D., et al. 1989, in Asteroids II, ed. R. P. Binzel, T. Gehrels, & M. S. Matthews, 524–556

    1989

  7. [7]

    2025, astropy/photutils: 2.2.0, 2.2.0 Zenodo, doi: 10.5281/zenodo.14889440

    Bradley, L., Sip˝ ocz, B., Robitaille, T., et al. 2025, astropy/photutils: 2.2.0, 2.2.0 Zenodo, doi: 10.5281/zenodo.14889440

    doi:10.5281/zenodo.14889440 2025

  8. [8]

    R., & Greisen, E

    Calabretta, M. R., & Greisen, E. W. 2002, A&A, 395, 1077, doi: 10.1051/0004-6361:20021327

    doi:10.1051/0004-6361:20021327 2002

  9. [9]

    C., Magnier, E

    Chambers, K. C., Magnier, E. A., Metcalfe, N., et al. 2016, arXiv e-prints. https://arxiv.org/abs/1612.05560

    Pith/arXiv arXiv 2016

  10. [10]

    2020, in AAS/Division for Planetary Sciences Meeting Abstracts, Vol

    Eggl, S., Juric, M., Moeyens, J., & Jones, L. 2020, in AAS/Division for Planetary Sciences Meeting Abstracts, Vol. 52, AAS/Division for Planetary Sciences Meeting Abstracts #52, 211.01

    2020

  11. [11]

    R., & Micheli, M

    Farnocchia, D., Chesley, S. R., & Micheli, M. 2015, Icarus, 258, 18, doi: 10.1016/j.icarus.2015.06.029

    work page doi:10.1016/j.icarus.2015.06.029 2015

  12. [12]

    2025, Multi-year stacking searches for solar system bodies, https://arxiv.org/abs/2509.25428

    Geringer-Sameth, A., Golovich, N., & Iwabuchi, K. 2025, Multi-year stacking searches for solar system bodies, https://arxiv.org/abs/2509.25428

    arXiv 2025

  13. [13]

    M., & Walker, M

    Geringer-Sameth, A., Koushiappas, S. M., & Walker, M. G. 2015, PhRvD, 91, 083535, doi: 10.1103/PhysRevD.91.083535

    doi:10.1103/physrevd.91.083535 2015

  14. [14]

    D., Yeomans, D

    Giorgini, J. D., Yeomans, D. K., Chamberlin, A. B., et al. 1996, in AAS/Division for Planetary Sciences Meeting

    1996

  15. [15]

    2025, Astronomy and Computing, 53, 100987, doi: 10.1016/j.ascom.2025.100987

    Golovich, N., Steil, T., Geringer-Sameth, A., et al. 2025, Astronomy and Computing, 53, 100987, doi: 10.1016/j.ascom.2025.100987

    arXiv 2025

  16. [16]

    R., Millman, K

    Harris, C. R., Millman, K. J., van der Walt, S. J., et al. 2020, Nature, 585, 357, doi: 10.1038/s41586-020-2649-2

    doi:10.1038/s41586-020-2649-2 2020

  17. [17]

    2024, in AAS/Division of Dynamical Astronomy Meeting, Vol

    Holman, M., & Rein, H. 2024, in AAS/Division of Dynamical Astronomy Meeting, Vol. 55, AAS/Division of Dynamical Astronomy Meeting, 503.03

    2024

  18. [18]

    2018, AJ, 156, 135, doi: 10.3847/1538-3881/aad69a

    Kuindersma, S. 2018, AJ, 156, 135, doi: 10.3847/1538-3881/aad69a

    doi:10.3847/1538-3881/aad69a 2018

  19. [19]

    Huber, P. J. 1964, The Annals of Mathematical Statistics, 35, 73 , doi: 10.1214/aoms/1177703732

    arXiv 1964

  20. [20]

    Hunter, J. D. 2007, Computing in Science & Engineering, 9, 90, doi: 10.1109/MCSE.2007.55 14 IRSA. 2022, Zwicky Transient Facility Image Service, IPAC, doi: 10.26131/IRSA539 Ivezi´ c,ˇZ., Kahn, S. M., Tyson, J. A., et al. 2019, The Astrophysical Journal, 873, 111, doi: 10.3847/1538-4357/ab042c

    doi:10.1109/mcse.2007.55 2007

  21. [21]

    S., & Li, J

    Jewitt, D., Stuart, J. S., & Li, J. 2011, AJ, 142, 28, doi: 10.1088/0004-6256/142/1/28

    doi:10.1088/0004-6256/142/1/28 2011

  22. [22]

    King, I. R. 1983, Publications of the Astronomical Society of the Pacific, 95, 163, doi: 10.1086/131139

    doi:10.1086/131139 1983

  23. [23]

    2024, in European Planetary Science Congress, EPSC2024–882, doi: 10.5194/epsc2024-882

    Licandro, J., Conversi, L., Delbo, M., et al. 2024, in European Planetary Science Congress, EPSC2024–882, doi: 10.5194/epsc2024-882

    work page doi:10.5194/epsc2024-882 2024

  24. [24]

    1936, Proceedings of the National Institute of Sciences of India, 2, 49

    Mahalanobis, P. 1936, Proceedings of the National Institute of Sciences of India, 2, 49

    1936

  25. [25]

    K., Masiero, J

    Mainzer, A. K., Masiero, J. R., Abell, P. A., et al. 2023, PSJ, 4, 224, doi: 10.3847/PSJ/ad0468

    doi:10.3847/psj/ad0468 2023

  26. [26]

    J., Laher, R

    Masci, F. J., Laher, R. R., Rusholme, B., et al. 2019, Publications of the Astronomical Society of the Pacific, 131, 018003, doi: 10.1088/1538-3873/aae8ac

    doi:10.1088/1538-3873/aae8ac 2019

  27. [27]

    R., & Sansaturio, M

    Milani, A., Chesley, S. R., & Sansaturio, M. E. 2005, Icarus, 173, 362, doi: 10.1016/j.icarus.2004.09.002

    doi:10.1016/j.icarus.2004.09.002 2005

  28. [28]

    Milani, A., & Gronchi, G. F. 2010, Theory of Orbit Determination (Cambridge University Press), doi: 10.1017/CBO9781139175371

    work page doi:10.1017/cbo9781139175371 2010

  29. [29]

    P., Micheli, M., Farnocchia, D., et al

    Naidu, S. P., Micheli, M., Farnocchia, D., et al. 2021, ApJL, 913, L6, doi: 10.3847/2041-8213/abf836

    work page doi:10.3847/2041-8213/abf836 2021

  30. [30]

    S., Folkner, W

    Park, R. S., Folkner, W. M., Williams, J. G., & Boggs, D. H. 2021, The Astronomical Journal, 161, 105, doi: 10.3847/1538-3881/abd414

    doi:10.3847/1538-3881/abd414 2021

  31. [31]

    Rein, H., & Spiegel, D. S. 2015, MNRAS, 446, 1424, doi: 10.1093/mnras/stu2164

    doi:10.1093/mnras/stu2164 2015

  32. [32]

    2013, in Highlights of Spanish Astrophysics VII, ed

    Rodrigo, C., Solano, E., & Pulido, R. 2013, in Highlights of Spanish Astrophysics VII, ed. J. C. Guirado, L. M. Lara, V. Quilis, & J. Gorgas, 995–1000

    2013

  33. [33]

    2023, A&A, 673, A93, doi: 10.1051/0004-6361/202345957

    Saifollahi, T., Verdoes Kleijn, G., Williams, R., et al. 2023, A&A, 673, A93, doi: 10.1051/0004-6361/202345957

    work page doi:10.1051/0004-6361/202345957 2023

  34. [34]

    2014, Astronomische Nachrichten, 335, 142, doi: 10.1002/asna.201311888

    Solano, E., Rodrigo, C., Pulido, R., & Carry, B. 2014, Astronomische Nachrichten, 335, 142, doi: 10.1002/asna.201311888

    doi:10.1002/asna.201311888 2014

  35. [35]

    Stetson, P. B. 1987, PASP, 99, 191, doi: 10.1086/131977

    doi:10.1086/131977 1987

  36. [36]

    H., et al

    Stetzler, S., Juri´ c, M., Bernardinelli, P. H., et al. 2025, arXiv e-prints, arXiv:2509.26279, doi: 10.48550/arXiv.2509.26279

    work page doi:10.48550/arxiv.2509.26279 2025

  37. [37]

    2009, Astronomische Nachrichten, 330, 698, doi: 10.1002/asna.200811198

    Vaduvescu, O., Curelaru, L., Birlan, M., et al. 2009, Astronomische Nachrichten, 330, 698, doi: 10.1002/asna.200811198

    work page doi:10.1002/asna.200811198 2009

  38. [38]

    2013, Astronomische Nachrichten, 334, 718, doi: 10.1002/asna.201211720

    Vaduvescu, O., Popescu, M., Comsa, I., et al. 2013, Astronomische Nachrichten, 334, 718, doi: 10.1002/asna.201211720

    work page doi:10.1002/asna.201211720 2013

  39. [39]

    E., & Hestroffer, D

    Vavilov, D. E., & Hestroffer, D. 2024, A&A, 689, A49, doi: 10.1051/0004-6361/202449830

    work page doi:10.1051/0004-6361/202449830 2024

  40. [40]

    2001, Icarus, 154, 412, doi: 10.1006/icar.2001.6592

    Virtanen, J., Muinonen, K., & Bowell, E. 2001, Icarus, 154, 412, doi: 10.1006/icar.2001.6592

    arXiv 2001

  41. [41]

    E., et al

    Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature Methods, 17, 261, doi: 10.1038/s41592-019-0686-2

    doi:10.1038/s41592-019-0686-2 2020