Pith. sign in

REVIEW 3 major objections 5 minor 66 references

HOLISMOKES XV. Search for strong gravitational lenses combining ground-based and space-based imaging

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Combining HST images with Legacy Survey images through a merged-branch neural network detects more gravitational lenses at a fixed false-positive rate than using either data source alone.

desk verdict A clean architectural comparison on simulated lenses — the merged HST+LS branch wins — but the headline TPRs are measured on mocks from the same pipeline, so the claim about real survey data is not yet established. read the letter →

arxiv 2411.18694 v1 pith:7XHNPRJQ submitted 2024-11-27 astro-ph.IM astro-ph.CO

A. Melo , R. Cañameras , S. Schuldt , S. H. Suyu , Irham T. Andika , S. Bag , S. Taubenberger This is my paper · ORCID
classification astro-ph.IMastro-ph.CO
keywords stronggravitationallensinggalaxy-scalelensesconvolutionalneuralnetworksmulti-resolutionimagefusionEuclidLSSTmocksimulationsResNet-18
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

Strong gravitational lenses are rare, and the coming wide surveys Euclid and LSST will produce far more galaxy images than humans can inspect, so automated classifiers must recover as many lenses as possible for a fixed number of false alarms. This paper argues that the best classifier uses two complementary views at once: sharp single-band space imaging from HST, which resolves arcs and Einstein rings, and blurrier three-band ground imaging from the Legacy Survey, which adds color. The authors insert simulated lenses into real galaxy images and train four ResNet-18 classifiers, comparing them at a false-positive rate of $10^{-4}$. The merged-branch network, which processes HST and LS through separate branches and fuses their features before classification, reaches a true-positive rate near 0.55, versus about 0.41 for HST alone, 0.45 for the ground data alone, and 0.51 for a stacked four-band variant. If the ranking carries over to Euclid and LSST, surveys expected to yield roughly $10^5$ lenses would harvest a meaningfully larger share of them at no extra observing cost.

What carries the argument

The load-bearing mechanism is the merged-branches ResNet-18 architecture. One branch ingests the $200\times200$ HST F814W cutout and the other ingests the $38\times38$ three-band LS cutout, each through residual convolutional blocks; the flattened 512-feature vectors are concatenated into a 1024-dimensional vector, reduced by a fully connected layer, and classified. This preserves each instrument's native pixel scale instead of forcing all data onto a common grid. The training set is built by simulating singular isothermal ellipsoid lenses with GLEE, painting arcs from Hubble Ultra Deep Field sources onto real luminous red galaxy images in both HST and LS with identical geometry, keeping only systems whose arcs are detected above $2\sigma$ in all three LS bands, and rotating the images to reach roughly 24,000 positive examples.

What would settle it

Apply the four trained networks to a sample of real, spectroscopically confirmed gravitational lenses and an equal number of independent non-lens galaxies from Euclid and LSST overlap fields, and check whether the merged-branch true-positive rate at a false-positive rate of $10^{-4}$ is measurably above the HST-only and LS-only rates.

Watch

Extended reading notes

Core claim

The paper's central claim is that fusing high-resolution single-band and low-resolution multi-band images with a two-branch convolutional network detects more galaxy-scale strong lenses than either data source alone, and more than a naive stack of all bands on a common pixel grid. On a test set of roughly 8,000 simulated lenses plus 364 real lens candidates, at a false-positive rate of $10^{-4}$ the true-positive rate is about 0.41 for HST F814W alone, 0.45 for Legacy Survey $g/r/z$ alone, 0.51 when the LS bands are interpolated to HST pixel scale and processed as a four-channel stack, and 0.55 when HST and LS pass through separate ResNet-18 branches whose 512-dimensional feature vectors are concatenated before classification. The merged architecture assigns its highest-confidence scores to small-Einstein-radius systems ($\theta_{\rm E} \le 0.8''$), the regime where single-instrument searches have historically struggled. A control experiment that downsampled HST to LS resolution and treated it as a fourth ground band performed like the stacked model, showing that the gain comes from preserving native resolution in its own branch.

Load-bearing premise

The mock lenses, simulated Einstein arcs pasted onto real galaxy images and kept only when visible above $2\sigma$ in all three ground-based bands, represent the real galaxy-scale lens population that Euclid and LSST will observe, so the measured ranking of architectures transfers to survey data.

Editorial extensions

If this is right

  • Applying the merged-branch design to Euclid VIS plus LSST $g/r/z$ should recover more galaxy-scale lenses at a fixed false-positive budget than either survey alone.
  • A true-positive rate near 0.55 at $10^{-4}$ false positives, projected onto surveys expected to contain roughly 100,000 lenses, implies tens of thousands of additional recovered systems compared with single-instrument classifiers.
  • Because the merged network scores low-Einstein-radius systems ($\theta_{\rm E} \le 0.8''$) with high confidence, small-separation lenses become more accessible to automated discovery.
  • The same architecture can fuse Euclid's own VIS and NISP channels, extending multi-resolution fusion beyond the HST/LS combination tested here.
  • For transient follow-up programs hunting lensed supernovae, a higher true-positive rate at low false-positive rate reduces the human inspection bottleneck.

Reading between the lines

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

  • Editorial inference: the merged-branch advantage suggests a general design rule that each instrument's native pixel scale should be processed in its own network branch and fused late, rather than interpolating all inputs onto one grid.
  • Editorial inference: the paper's own caveat that negative cutouts come from the same fields as the lenses means the reported false-positive rates could be optimistic, and a cleaner test would build negatives from fully independent sky regions.
  • Editorial inference: a testable extension is to measure the merged-branch ROC on real confirmed lenses in Euclid Early Release Observations overlapping LSST coverage, since the gain over single-instrument models could shrink when the training mocks no longer match the data.
  • Editorial inference: because the mock sample only keeps arcs visible above $2\sigma$ in all three LS bands, the measured fusion gain may not apply to the very faintest arcs that only HST can see, and restricting the test set to faint arcs would bound the effect.
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 / 5 minor

Summary. The paper proposes and evaluates four ResNet-18-based CNN architectures for detecting galaxy-scale strong gravitational lenses by combining high-resolution single-band HST F814W images with lower-resolution, multi-band Legacy Survey g/r/z images. The four architectures are LS-only, HST-only, stacked (LS interpolated to HST pixel scale and processed together), and merged (separate HST and LS branches fused before the fully connected layer). Training and test data are dominated by mock lenses simulated with the GLEE pipeline using LRG deflectors and HUDF source images; only 364 real lenses are included in the test set and are not scored separately in the ROC analysis. At a false-positive rate of 10^-4, the authors report TPRs of about 0.41, 0.45, 0.51, and 0.55 for the four architectures and conclude that multi-instrument fusion significantly enhances detection. The paper also includes a control with downsampled HST images as a fourth LS filter, which supports the claim that merging is not equivalent to stacking.

Significance. If the reported gains hold on real survey data, the merged-branch architecture would be a useful template for combined Euclid and LSST lens searches, and the controlled comparison of fusion strategies is methodologically valuable. The downsampled-HST control and the use of a large negative set for low-FPR evaluation are strengths. However, the central quantitative claim rests on a test set that is about 96% simulated with the same pipeline used for training, and the real-lens subset is never evaluated separately. The external validity of the headline TPR improvement is therefore not yet established, although the issue is addressable within the scope of the manuscript.

major comments (3)
  1. [Section 4, Table 2, Fig. 10] The headline TPR comparison is computed on a positive test set containing roughly 8,000 mocks and only 364 real lenses. Because the mocks are generated with the same GLEE pipeline and HUDF source library used for training, and are selected (Sect. 2, Fig. 5) to have arcs visible at >2 sigma in all three LS bands, the test largely measures the networks' ability to recognize simulation-pipeline lenses rather than real galaxy-scale lenses. The paper does not report a separate ROC, TPR, or score statistics for the 364 real positives, which appear only as individual examples in Fig. 13. Please add this analysis and explicitly discuss how the 2-sigma LS-arc-visibility selection conditions the measured gain; without it, the claim that merged HST+LS branches enhance detection for real survey data is not supported.
  2. [Section 4, Table 2] No error bars or uncertainty statements are provided for the TPR values at FPR=10^-4. At that FPR with about 120,000 negatives, the operating point is defined by only about 12 false positives, so the estimates are subject to non-negligible Poisson and threshold-sampling noise. The ten-split comparison in Fig. 10 (right) addresses only stacked versus merged and does not quantify the significance of the difference between single-instrument and fused models, which is the central 'significantly enhance' claim. Please propagate these uncertainties and, if possible, report bootstrap or split-half confidence intervals.
  3. [Abstract, Section 4, Table 2] The reported TPR ordering for HST and LS is internally inconsistent: the abstract and Section 4 text state HST ~0.41 and LS ~0.45, while Table 2 lists LS=0.41 and HST=0.45. Because these numbers are the basis for the architecture ranking, this inconsistency must be resolved and the corrected values verified throughout the manuscript.
minor comments (5)
  1. [Section 2, Table 1] Table 1 lists positive counts summing to about 25,900, while the text states that rotation yields approximately 24,000 positive images and that the test set contains about 8,000 mocks plus 364 real lenses; please reconcile these numbers.
  2. [Section 3] The description of balancing 'LS and HST contribute equally to the overall loss function' is unclear, since in the merged architecture the loss is computed after the fused feature vector; please specify the exact weighting scheme.
  3. [Section 4, Fig. 10] The phrase 'ten different test sets' is not defined precisely; please state whether the models were retrained for each split, how the splits were drawn, and whether real lenses were included in each split.
  4. [Section 4] The sentence claiming that 'the estimates from HST and LS alone are less reliable in the low-FPR regime compared to the combined data set' is unexplained, because all models are evaluated on the same negative set; please justify or remove this statement.
  5. [Section 2 and Fig. 13 caption] There are minor typographical errors, including 'simulated simulated HST' in Section 2 and a duplicated article in the Fig. 13 caption ('compared to the the LS-only model').

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the ROC comparison is an empirical evaluation on a held-out test set, not a derivation from its inputs.

full rationale

The paper makes no first-principles derivation; its claims are empirical ROC comparisons on a held-out test set. The test set is disjoint from training, the four architectures share identical splits, and the ROC thresholds are evaluated on test outputs rather than fitted to test labels. The dominant presence of mock positives generated with the same GLEE/HUDF pipeline used for training is a representativeness/domain-shift limitation: the quoted TPRs may not transfer to real survey data, but this is not a reduction of the conclusion to the inputs. Similarly, the 2-sigma LS-visibility selection restricts the evaluation regime without making the merged-branch advantage true by construction. The self-citations (Schuldt et al. 2021 simulation recipe; Cañameras et al. 2021 ResNet-18; Rojas et al. 2022 KNN) are ordinary method citations, not load-bearing uniqueness claims. The inconsistency between Table 2 and the Abstract regarding which of LS/HST achieves TPR 0.41 vs 0.45 is a correctness flag, not circularity. Under the provided definitions, no equation or fitted parameter reduces to its own input, so the appropriate finding is no significant circularity.

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

The central claim rests on a set of hand-chosen simulation and selection parameters, plus four domain assumptions about the transferability of mock training data to real surveys. The paper does not release code or mock catalogs, so these choices are not externally auditable. No new physical entities are introduced.

free parameters (4)
  • Einstein radius selection range = 0.5-2.5 arcsec
    Hand-chosen range (Sect. 2) that defines the simulated test population; TPR is measured over this range.
  • 2-sigma arc visibility threshold = >2 sigma in HST and all three LS bands
    Hand-chosen cut (Sect. 2) that removes faint arcs from the mock test set, biasing the sample toward bright, detectable configurations and potentially inflating TPR.
  • KNN velocity dispersion prediction = rms scatter 65 km/s
    The lens velocity dispersions for the mocks are predicted by a KNN trained on 580k SDSS/LS LRGs (Sect. 2); the mock population inherits the scatter of this predictor.
  • Rotational augmentation factor = 4 (0, 90, 180, 270 degrees)
    The training set is inflated by rotating the LRG images and simulating new lenses; the effective sample size and its diversity depend on this choice.
assumptions (4)
  • domain assumption SIE mass profile is an adequate model for galaxy-scale lens deflectors in the simulated sample.
    The mocks use a Singular Isothermal Ellipsoid model (Sect. 2); if real deflectors have different mass distributions, mock-based training and testing may not transfer.
  • domain assumption HST F814W and Legacy Survey g,r,z are adequate proxies for Euclid VIS and LSST bands.
    The paper motivates the setup by comparing filters (Fig. 1) and expects 'even better performance' with Euclid (Sect. 5), but the actual transfer is not tested.
  • domain assumption The redshift and spectral energy distributions of the HUDF sources are representative of the background sources that will be lensed in Euclid/LSST.
    Source images are drawn from the HUDF catalog of Schuldt et al. (2021) with z in 0.1-4.0 (Sect. 2).
  • domain assumption Non-lens contaminants in the test set are representative of the false-positive population in Euclid/LSST.
    The authors acknowledge using LRGs from the same simulations as negatives can create repetitions between training and test sets (Sect. 2), so the negative population is not independent.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HOLISMOKES XV. Search for strong gravitational lenses combining ground-based and space-based imaging." pith.science (2026). https://pith.science/paper/7XHNPRJQ

@misc{pith2026241118694,
  author       = {Pith},
  title        = {Pith review of: HOLISMOKES XV. Search for strong gravitational lenses combining ground-based and space-based imaging},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7XHNPRJQ}},
  note         = {Machine review of arXiv:2411.18694}
}
abstract

In the past, researchers have mostly relied on single-resolution images from individual telescopes to detect gravitational lenses. We propose a search for galaxy-scale lenses that, for the first time, combines high-resolution single-band images (in our case the Hubble Space Telescope, HST) with lower-resolution multi-band images (in our case Legacy survey, LS) using machine learning. This methodology aims to simulate the operational strategies that will be employed by future missions, such as combining the images of Euclid and the Rubin Observatory's Legacy Survey of Space and Time (LSST). To compensate for the scarcity of lensed galaxy images for network training, we have generated mock lenses by superimposing arc features onto HST images, saved the lens parameters, and replicated the lens system in the LS images. We test four architectures based on ResNet-18: (1) using single-band HST images, (2) using three bands of LS images, (3) stacking these images after interpolating the LS images to HST pixel scale for simultaneous processing, and (4) merging a ResNet branch of HST with a ResNet branch of LS before the fully connected layer. We compare these architecture performances by creating Receiver Operating Characteristic (ROC) curves for each model and comparing their output scores. At a false-positive rate of $10^{-4}$, the true-positive rate is $\sim$0.41, $\sim$0.45, $\sim$0.51 and $\sim$0.55, for HST, LS, stacked images and merged branches, respectively. Our results demonstrate that models integrating images from both the HST and LS significantly enhance the detection of galaxy-scale lenses compared to models relying on data from a single instrument. These results show the potential benefits of using both Euclid and LSST images, as wide-field imaging surveys are expected to discover approximately 100,000 lenses.

Figures

Figures reproduced from arXiv: 2411.18694 by the authors.

Figure 1
Figure 1. Some of the filters from LSST (in shaded color labeled [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Galaxy parameters (zd, zs , and vdisp) of the lenses and sources used for simulating galaxy-scale lenses. (Le f t) Histogram of the number of galaxies as a function of the redshift of the source zs (red) and lens zd (blue). (Right) Normalized vdisp distribution measured from SDSS used for training the KNN network (orange color), and the predicted normalized vdisp distribution for lens galaxies over the final sample … view at source ↗
Figure 3
Figure 3. Distribution of the Einstein radii of the mock strong lens [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Simulation pipeline procedure used for simulating galaxy-scale lenses. See further details in Sec. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Properties of the 10,129 HST and LS lens simulations: ( [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Positive lens simulations of LS and HST for di [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Visualized architectures for the four scenarios: (a) LS, (b) HST architecture, (c) stacked-images architecture, and (d) merged [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Loss curves for the training (blue and light blue colors) and validation (purple and pink colors) for the four scenarios. Left: [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Learning curves for the stacked-images architecture using [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: ROC curves of different network architectures. Left panel: our four main networks illustrated in [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]
Figure 11
Figure 11. Figure 11: CNN scores for the positive examples of the test set. [PITH_FULL_IMAGE:figures/full_fig_p010_11.png]
Figure 12
Figure 12. Figure 12: CNN scores for systems with θE ≤ 0.8 ′′ from the test set scored by both architectures. introduced by varying resolutions. This limitation highlights the need for further investigation into how the architecture processes and combines multi-resolution data to improve i…
Figure 13
Figure 13. Figure 13: CNN scores of some test set examples of gravitational lensing ( [PITH_FULL_IMAGE:figures/full_fig_p011_13.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

66 extracted references · 24 canonical work pages

  1. [1]

    A., O’Riordan, C

    Acevedo Barroso, J. A., O’Riordan, C. M., Clément, B., et al. 2024, arXiv e- prints, arXiv:2408.06217

    arXiv 2024

  2. [2]

    2024, MNRAS, 531, 3509

    Arendse, N., Dhawan, S., Sagués Carracedo, A., et al. 2024, MNRAS, 531, 3509

    work page 2024

  3. [3]

    W., Treu, T., Bolton, A

    Auger, M. W., Treu, T., Bolton, A. S., et al. 2009, ApJ, 705, 1099

    2009

  4. [4]

    H., et al

    Bag, S., Huber, S., Suyu, S. H., et al. 2024, A&A, 691, A100

    work page 2024

  5. [5]

    Beckwith, S. V . W., Stiavelli, M., Koekemoer, A. M., et al. 2006, AJ, 132, 1729

    work page 2006

  6. [6]

    S., Burles, S., Koopmans, L

    Bolton, A. S., Burles, S., Koopmans, L. V . E., et al. 2008, ApJ, 682, 964

    2008

  7. [7]

    S., Burles, S., Koopmans, L

    Bolton, A. S., Burles, S., Koopmans, L. V . E., Treu, T., & Moustakas, L. A. 2006, ApJ, 638, 703

    2006

  8. [8]

    R., Bolton, A

    Brownstein, J. R., Bolton, A. S., Schlegel, D. J., et al. 2012, ApJ, 744, 41 Cañameras, R., Schuldt, S., Shu, Y ., et al. 2021, A&A, 653, L6 Cañameras, R., Schuldt, S., Suyu, S. H., et al. 2020, A&A, 644, A163

    work page 2012

Show all 66 references
  1. [9]

    2023, arXiv e-prints, arXiv:2306.03136

    Canameras, R., Schuldt, S., Shu, Y ., et al. 2023, arXiv e-prints, arXiv:2306.03136

    2023 arXiv

  2. [10]

    2007, ApJS, 172, 99

    Capak, P., Aussel, H., Ajiki, M., et al. 2007, ApJS, 172, 99

    2007

  3. [11]

    2019, arXiv e-prints, arXiv:1904.10439

    Capak, P., Cuillandre, J.-C., Bernardeau, F., et al. 2019, arXiv e-prints, arXiv:1904.10439

    2019 arXiv

  4. [12]

    Chan, J. H. H., Suyu, S. H., Sonnenfeld, A., et al. 2020, A&A, 636, A87

    2020

  5. [13]

    Chan, J. H. H., Wong, K. C., Ding, X., et al. 2024, MNRAS, 527, 6253

    2024

  6. [14]

    L., Oguri, M., et al

    Chen, W., Kelly, P. L., Oguri, M., et al. 2022, Nature, 611, 256

    2022

  7. [15]

    Collett, T. E. 2015, ApJ, 811, 20

    2015

  8. [16]

    T., Buckley-Geer, E

    Diehl, H. T., Buckley-Geer, E. J., Lindgren, K. A., et al. 2017, ApJS, 232, 15 Euclid Collaboration, Mellier, Y ., Abdurro’uf, et al. 2024, arXiv e-prints, arXiv:2405.13491

    2017

  9. [17]

    2008, ApJS, 176, 19

    Faure, C., Kneib, J.-P., Covone, G., et al. 2008, ApJS, 176, 19

    2008

  10. [18]

    L., Pascale, M., Pierel, J., et al

    Frye, B. L., Pascale, M., Pierel, J., et al. 2024, ApJ, 961, 171

    2024

  11. [19]

    O., Kruk, S., Cornen, C., et al

    Garvin, E. O., Kruk, S., Cornen, C., et al. 2022, A&A, 667, A141

    2022

  12. [20]

    A., Nugent, P

    Goldstein, D. A., Nugent, P. E., & Goobar, A. 2019, ApJS, 243, 6

    2019

  13. [21]

    R., et al

    Goobar, A., Amanullah, R., Kulkarni, S. R., et al. 2017, Science, 356, 291

    2017

  14. [22]

    A., Johansson, J., Dhawan, S., et al

    Goobar, A. A., Johansson, J., Dhawan, S., et al. 2022, Transient Name Server AstroNote, 180, 1

    2022

  15. [23]

    2016, in 2016 IEEE Conference on Com- puter Vision and Pattern Recognition (CVPR, 1

    He, K., Zhang, X., Ren, S., & Sun, J. 2016, in 2016 IEEE Conference on Com- puter Vision and Pattern Recognition (CVPR, 1

    2016

  16. [24]

    2020, ApJ, 894, 78

    Huang, X., Storfer, C., Ravi, V ., et al. 2020, ApJ, 894, 78

    2020

  17. [25]

    2017, A&A, 608, A2

    Inami, H., Bacon, R., Brinchmann, J., et al. 2017, A&A, 608, A2

    2017

  18. [26]

    N., et al

    Ivezic, Z., Axelrod, T., Brandt, W. N., et al. 2008, Serbian Astronomical Journal, 176, 1 Ivezi´c, Ž., Kahn, S. M., Tyson, J. A., et al. 2019, ApJ, 873, 111

    2008

  19. [27]

    2019, ApJS, 243, 17

    Jacobs, C., Collett, T., Glazebrook, K., et al. 2019, ApJS, 243, 17

    2019

  20. [28]

    T., More, A., Oguri, M., et al

    Jaelani, A. T., More, A., Oguri, M., et al. 2020, MNRAS, 495, 1291

    2020

  21. [29]

    T., More, A., Wong, K

    Jaelani, A. T., More, A., Wong, K. C., et al. 2023, arXiv e-prints, arXiv:2312.07333

    2023 arXiv

  22. [30]

    & Kovner, I

    Kassiola, A. & Kovner, I. 1993, ApJ, 417, 450

    1993

  23. [31]

    2022, Transient Name Server AstroNote, 169, 1

    Kelly, P., Zitrin, A., Oguri, M., et al. 2022, Transient Name Server AstroNote, 169, 1

    2022

  24. [32]

    L., Rodney, S

    Kelly, P. L., Rodney, S. A., Treu, T., et al. 2015, Science, 347, 1123

    2015

  25. [33]

    1995, in Astronomical Society of the Pacific Conference Series, V ol

    Krist, J. 1995, in Astronomical Society of the Pacific Conference Series, V ol. 77, Astronomical Data Analysis Software and Systems IV , ed. R. A. Shaw, H. E. Payne, & J. J. E. Hayes, 349

    1995

  26. [34]

    2018, MNRAS, 473, 3895

    Lanusse, F., Ma, Q., Li, N., et al. 2018, MNRAS, 473, 3895

    2018

  27. [35]

    2011, arXiv e-prints, arXiv:1110.3193

    Laureijs, R., Amiaux, J., Arduini, S., et al. 2011, arXiv e-prints, arXiv:1110.3193

    2011 arXiv

  28. [36]

    B., Meneghetti, M., Avestruz, C., et al

    Metcalf, R. B., Meneghetti, M., Avestruz, C., et al. 2019, A&A, 625, A119

    2019

  29. [37]

    T., Jackson, N

    Myers, S. T., Jackson, N. J., Browne, I. W. A., et al. 2003, MNRAS, 341, 1

    2003

  30. [38]

    Nikutta, R., Fitzpatrick, M., Scott, A., & Weaver, B. A. 2020, Astronomy and Computing, 33, 100411

    2020

  31. [39]

    2006, AJ, 132, 999

    Oguri, M., Inada, N., Pindor, B., et al. 2006, AJ, 132, 999

    2006

  32. [40]

    & Marshall, P

    Oguri, M. & Marshall, P. J. 2010, MNRAS, 405, 2579

    2010

  33. [41]

    S., Courbin, F., Faure, C., Kokotanekova, R., & Meylan, G

    Pawase, R. S., Courbin, F., Faure, C., Kokotanekova, R., & Meylan, G. 2014, MNRAS, 439, 3392

    2014

  34. [42]

    E., Tortora, C., Vernardos, G., et al

    Petrillo, C. E., Tortora, C., Vernardos, G., et al. 2019, MNRAS, 484, 3879

    2019

  35. [43]

    Pierel, J. D. R., Newman, A. B., Dhawan, S., et al. 2024, ApJ, 967, L37

    2024

  36. [44]

    2023, A&A, 675, L4

    Polletta, M., Nonino, M., Frye, B., et al. 2023, A&A, 675, L4

    2023

  37. [45]

    2018, ApJ, 856, 68

    Pourrahmani, M., Nayyeri, H., & Cooray, A. 2018, ApJ, 856, 68

    2018

  38. [46]

    1964, MNRAS, 128, 307

    Refsdal, S. 1964, MNRAS, 128, 307

    1964

  39. [47]

    A., Brammer, G

    Rodney, S. A., Brammer, G. B., Pierel, J. D. R., et al. 2021, Nature Astronomy, 5, 1118

    2021

  40. [48]

    2022, A&A, 668, A73 Sainz de Murieta, A., Collett, T

    Rojas, K., Savary, E., Clément, B., et al. 2022, A&A, 668, A73 Sainz de Murieta, A., Collett, T. E., Magee, M. R., et al. 2024, arXiv e-prints, arXiv:2407.04080

    2022 arXiv

  41. [49]

    Schaefer, C., Geiger, M., Kuntzer, T., & Kneib, J. P. 2018, A&A, 611, A2

    2018

  42. [50]

    Schneider, P., Ehlers, J., & Falco, E. E. 1992, Gravitational Lenses

    1992

  43. [51]

    T., et al

    Schuldt, S., Canameras, R., Andika, I. T., et al. 2024, arXiv e-prints, arXiv:2405.20383

    2024 arXiv

  44. [52]

    H., Meinhardt, T., et al

    Schuldt, S., Suyu, S. H., Meinhardt, T., et al. 2021, A&A, 646, A126

    2021

  45. [53]

    G., Aussel, H., et al

    Scoville, N., Abraham, R. G., Aussel, H., et al. 2007, ApJS, 172, 38

    2007

  46. [54]

    2022, A&A, 662, A4

    Shu, Y ., Cañameras, R., Schuldt, S., et al. 2022, A&A, 662, A4

    2022

  47. [55]

    Sonnenfeld, A., Chan, J. H. H., Shu, Y ., et al. 2018, PASJ, 70, S29

    2018

  48. [56]

    Suyu, S. H. & Halkola, A. 2010, A&A, 524, A94

    2010

  49. [57]

    H., Hensel, S

    Suyu, S. H., Hensel, S. W., McKean, J. P., et al. 2012, ApJ, 750, 10

    2012

  50. [58]

    H., Huber, S., Cañameras, R., et al

    Suyu, S. H., Huber, S., Cañameras, R., et al. 2020, A&A, 644, A162

    2020

  51. [59]

    L., Ostriker, J

    Turner, E. L., Ostriker, J. P., & Gott, J. R., I. 1984, ApJ, 284, 1

    1984

  52. [60]

    2022, MNRAS, 509, 3966

    Walmsley, M., Lintott, C., Géron, T., et al. 2022, MNRAS, 509, 3966

    2022

  53. [61]

    M., et al

    Wilde, J., Serjeant, S., Bromley, J. M., et al. 2022, MNRAS, 512, 3464

    2022

  54. [62]

    W., Galloway, M

    Willett, K. W., Galloway, M. A., Bamford, S. P., et al. 2017, MNRAS, 464, 4176

    2017

  55. [63]

    2019, MNRAS, 487, 3342

    Wojtak, R., Hjorth, J., & Gall, C. 2019, MNRAS, 487, 3342

    2019

  56. [64]

    C., Chan, J

    Wong, K. C., Chan, J. H. H., Chao, D. C. Y ., et al. 2022, PASJ, 74, 1209

    2022

  57. [65]

    L., Eisenhardt, P

    Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868

    2010

  58. [66]

    A., et al

    Zhou, R., Dey, B., Newman, J. A., et al. 2023, AJ, 165, 58 Article number, page 12 of 12

    2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.