Pith. sign in

REVIEW 4 major objections 5 minor 27 references

Tools for High Precision Photometry from Wide-Field Color Images

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read AstroWISP extracts few-parts-per-thousand photometry from Bayer-masked DSLR images, enough to detect exoplanet transits.

desk verdict A genuinely useful, packaged Bayer-mask photometry tool with a real transit detection, but the outperformance claim and per-exposure precision numbers need direct comparison and more careful systematics handling. read the letter →

arxiv 2507.14289 v2 pith:HBYFWJA3 submitted 2025-07-18 astro-ph.IM astro-ph.SR

Kaloyan Penev , Angel Romero , S. Javad Jafarzadeh , Olivier Guyon , Wilfred Gee , Preethi Krishnamoorthy This is my paper · ORCID
classification astro-ph.IMastro-ph.SR
keywords BayermaskaperturephotometryPSFfittingDSLRcamerasexoplanettransitscitizensciencesubpixelsensitivitywide-field
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper presents AstroWISP, a set of image-processing tools for source extraction, background determination, PSF/PRF fitting, and aperture photometry, aimed at the color detectors found in consumer DSLR cameras. The authors claim that AstroWISP reaches a few parts per thousand photometric precision per 2-minute exposure on real PANOPTES data from Bayer-masked cameras, significantly outperforming prior custom efforts. They demonstrate that the tools recover the WASP-33b transit with a depth of $16.51\pm0.32$ ppt, matching the literature value of 16.78 ppt, even though the images were raw and uncalibrated. The purpose is to make precision time-series photometry feasible for citizen scientists using inexpensive off-the-shelf equipment.

What carries the argument

The load-bearing object is the per-pixel correction factor $k_p$ used in aperture photometry, defined as the ratio of the integral of the PSF-plus-background over the pixel-aperture intersection to the integral of the PSF times the Bayer sensitivity map over the full pixel. AstroWISP evaluates these integrals exactly, with machine precision, using closed-form expressions for integrals of arbitrary polynomials over all circle-rectangle overlaps. The PSF/PRF is represented as a piecewise bicubic function whose coefficients are set by values and mixed derivatives on interior cell corners, constrained to be smooth and to vanish on the outer grid boundary, with the source flux a separate global scaling factor. This machinery removes the two approximations that degrade standard photometry on color images: treating one color channel as a densely packed pixel grid and assuming a uniform flux distribution inside each pixel.

What would settle it

Take the same field and camera and reduce the images with and without flat-field calibration; if the few-parts-per-thousand scatter appears only in the uncalibrated case, the claim that such precision is achievable from raw data would not generalize. Alternatively, apply a dithered sequence and shift the assumed Bayer-mask registration by tenths of a pixel; if the measured flux of a constant star varies by more than the quoted precision, the assumed sensitivity map is not accurate enough.

Watch

Extended reading notes

Core claim

The central claim is that the Bayer mask, normally a serious obstacle to precision photometry, can be modeled exactly rather than approximated away. AstroWISP accounts for the nonuniform subpixel sensitivity pattern $S(x,y)$ by computing exact analytical integrals of polynomial PSF models over the overlap between circular apertures and rectangular pixels, and it propagates pixel noise formally to every reported flux. On raw, uncalibrated PANOPTES DSLR images, this yields per-exposure scatter of a few parts per thousand in combined-channel light curves and better than 1% in individual color channels, and it detects the WASP-33b transit with depth $16.51^{+0.32}_{-0.30}$ ppt, timing offset $6.59^{+1.37}_{-1.35}$ minutes from the literature ephemeris, and duration $2.56\pm0.05$ hours. The authors take this as evidence that consumer color cameras can be used for transit detection and ephemeris maintenance without calibration data, and that precision should improve further when calibration frames and fixed exposure times are used.

Load-bearing premise

The method assumes that the Bayer-mask sensitivity map $S(x,y)$ is known and static and that pixel-level effects such as charge bleeding, brighter-fatter, and antiblooming nonlinearity are negligible; the demonstration on raw, uncalibrated images does not independently check this.

Editorial extensions

If this is right

  • Amateur astronomers using consumer DSLR cameras can produce light curves precise enough to detect exoplanet transits, enabling community-based monitoring of long-period, bright-host planets.
  • The exact circle-rectangle integration applies beyond color detectors, improving photometry for monochrome cameras with nonuniform pixel response or undersampled PSFs.
  • The demonstrated work on raw, uncalibrated images lowers the practical burden of calibration data collection for citizen-science networks, while suggesting that adding darks and flats would push precision still lower.
  • The PSF/PRF fitting with smoothness constraints and external-catalog source positions can be reused in professional surveys that need careful handling of blended or poorly sampled sources.
  • Packaging the compiled tools as wheels for Windows, MacOS, and Linux makes the functionality accessible to a broad community without a compilation step.

Reading between the lines

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

  • If the few-parts-per-thousand precision holds across diverse camera models and site conditions, networks of amateur DSLR systems could complement professional surveys by watching bright stars with deep, long-period transits that TESS undersamples.
  • The $k_p$ formalism could be extended to absorb pixel crosstalk, brighter-fatter, and antiblooming nonlinearity by replacing the static sensitivity map with an intensity-dependent map, potentially lowering the precision floor below the current demonstration.
  • The exact integral technique could generalize to elliptical or otherwise non-circular apertures, which would matter for wide-field corners where PSFs are elongated, while preserving the same error-propagation scheme.
  • A direct testable extension is to compare flat-fielded and unflat-fielded light curves of constant stars from the same DSLR; the difference would show how much of the residual scatter is calibratable versus intrinsic to the Bayer-mask model.
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

4 major / 5 minor

Summary. The paper presents AstroWISP, an open-source C/C++ package with a Python wrapper for source extraction, background estimation, PSF/PRF fitting, and aperture photometry, aimed at high-precision photometry from consumer DSLR cameras with Bayer masks. The core algorithmic idea is to model subpixel sensitivity exactly by integrating polynomial PSF/PRF models over circle-rectangle overlaps, avoiding the approximation of treating color channels as densely packed pixels. The authors demonstrate the package on PANOPTES data: they detect the WASP-33b transit from uncalibrated DSLR images, measure a transit depth of 16.51^{+0.32}_{-0.30} ppt (literature 16.78 ppt), and report per-exposure median absolute deviations at the few-parts-per-thousand level for a field centered on FU Orionis. They claim this significantly outperforms prior custom photometry efforts for color images, and they package the tools for Windows, macOS, and Linux.

Significance. If the claims hold, AstroWISP is a valuable, well-packaged contribution to citizen-science photometry: it provides a cross-platform, freely available toolchain that models Bayer-mask effects without the need for custom per-project code, and the WASP-33b detection from raw, uncalibrated DSLR images is an encouraging empirical demonstration. The machine-checkable software distribution (Zenodo DOI, pre-built wheels, CI via cibuildwheel) and the use of real PANOPTES data are strengths. However, the headline claim of significantly outperforming prior efforts is not established by the evidence presented, and one algorithmic equation appears to contain an error. The contribution is therefore promising but requires revision.

major comments (4)
  1. [§3.2] The statement that AstroWISP enables 'a few parts per thousand photometric precision per exposure' and 'significantly outperforming prior efforts (e.g. Guyon & Martinache 2012; Zhang et al. 2016)' is not supported by the evidence presented. The only quantitative result is the median absolute deviation from the median of unbinned 2-min exposures for a single field; no side-by-side reduction of the same images with the cited prior methods is given, and the MAD scatter includes atmospheric, transparency, and detrending contributions that are not separated from the software's contribution. To support the claim, the authors should add a direct comparison using identical images and the same metric, or temper the claim to describe what is actually measured.
  2. [§3.1] The transit-depth agreement is reported as 16.51^{+0.32}_{-0.30} ppt versus a literature value of 16.78 ppt 'if we use our definition of the depth.' The depth is defined as the mean of the transit model over the central 50% of the time between first and fourth contact, which is not the standard full-transit depth reported in the exoplanet literature. The paper does not show how the 16.78 ppt literature value is converted to this definition, so the apparent agreement could be partly definitional. Please provide the literature-derived value evaluated with the same definition, or compare using the standard depth definition.
  3. [§2.2, Eq. (2)] The definition of k_p in Eq. (2) appears to have a technical error. The numerator integrates f+B over the pixel-aperture overlap without the subpixel sensitivity map S(x,y), while the denominator contains B plus an integral of f*S. If a pixel's sensitive area lies partly outside the aperture, this ratio does not give the fraction of the pixel's recorded flux that should be assigned to the aperture; for a Bayer pixel whose sensitive subarea is outside the aperture, k_p would be nonzero even though the pixel records no light from inside the aperture. The units also appear inconsistent because B is added to an integral of f over a dimensionless area. Please correct Eq. (2) or define all quantities with explicit units.
  4. [§2.1 and §2.2] The paper repeatedly states that AstroWISP uses 'exact analytical expressions' for integrals of arbitrary polynomials over all possible overlaps between circles and rectangles, ensuring machine precision without approximation, but it does not provide these expressions, a derivation, or a numerical validation. Since this exactness is a central algorithmic claim and distinguishes AstroWISP from approximate pixel-grid methods, please include the formulas in an appendix or provide a reference, together with a test showing agreement with high-resolution numerical integration.
minor comments (5)
  1. [Fig. 2 caption] The caption describes the small blue points as raw measurements but no per-point uncertainties are shown; please clarify whether the small points are individual color-channel measurements from all cameras, and describe how the binning and the RMS error bars were computed.
  2. [§3.1] The word 'egreess' in the sentence about ingress/egress should be 'egress'.
  3. [§1] The claim that 'none of these tools were designed with color detectors in mind' is strong; consider rewording to 'do not natively model Bayer-mask subpixel sensitivity' or provide a reference for a specific counterexample.
  4. [§2.1, Eq. (1)] The mapping from a point (x,y) to the cell indices (i,j) used in Eq. (1) is not explicitly defined; please state that i,j are chosen such that x_i < x-x_0 < x_{i+1} and y_j < y-y_0 < y_{j+1}.
  5. [§3.2] It would be helpful to state the number of stars in each magnitude bin and the number of exposures used for the median absolute deviation calculation, as well as any outlier rejection criteria.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: AstroWISP's photometric precision and transit-depth claims are empirical demonstrations checked against external data, not constructed predictions fitted to their own outputs.

full rationale

The paper's central claims are empirical demonstrations rather than derivations from assumed outputs. The WASP-33b transit depth and timing are fitted from PANOPTES DSLR images and compared to independent literature values and TESS data, so the result does not reduce to a fitted constant or to the paper's own model assumptions. The per-exposure photometric scatter in Sec. 3.2 is measured directly as the median absolute deviation of actual unbinned 2-minute exposures, not predicted from the PSF/PRF model. The Bayer-mask sensitivity map S(x,y) and the piecewise-polynomial PSF/PRF are model inputs, but no equation is constructed so that the claimed precision or transit depth is identical to those inputs by definition. The paper does cite its own companion software and papers, notably Romero et al. (2025) and the AstroWISP code release, but these citations are not load-bearing for the demonstrated photometric performance; the processing steps are described in the text and the results are externally benchmarked. The paper's own caveats, such as the lack of calibration data and the assumption that the sensitivity map is known and static, are limitations or correctness risks, not circularity. The comparative claim that AstroWISP significantly outperforms prior efforts is not supported by a direct side-by-side comparison, but that is an evidentiary gap rather than a circular reduction. Therefore no specific circular step can be exhibited, and the appropriate score is 0.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

No new physical entities are introduced; the super-pixel and Bayer-mask sensitivity map are standard descriptions of existing detectors. The free parameters are mostly internal modeling choices (PSF grid, background annulus, aperture radius) that are not reported in enough detail for exact reproduction.

free parameters (4)
  • PSF/PRF shape coefficients C_{i,j,m,n} = not reported
    Fitted by FitStarShape from images (Sec. 2.1, Eq. 1); the shape is the model input to all photometry and is not independently validated.
  • Background exclusion and outer radii = 6 and 13 pixels
    Chosen by hand for the PANOPTES camera and lens combination (Sec. 2.1); the authors note values may need adjustment, so results depend on this choice.
  • Aperture radius r = not reported
    SubPixPhot accepts a list of apertures (Sec. 2.2); final lightcurves must select or combine radii, and the specific radii used for Figs. 4 and 5 are not given.
  • PSF/PRF grid layout (M rows, N columns, nonuniform boundaries) = not reported
    The piecewise bicubic model is defined on a user-specified grid (Sec. 2.1); grid resolution affects how well the true PSF is represented.
assumptions (6)
  • domain assumption The PSF/PRF is representable as a piecewise bicubic polynomial with continuous first derivatives and zero values and derivatives at the outer grid boundary.
    Sec. 2.1, Eq. 1 and surrounding text; the model is flexible but not guaranteed to capture arbitrarily complex real PSFs, especially far from the field center or with strong color dependence.
  • domain assumption The pixel sensitivity map S(x,y) is known and static, and detector effects such as brighter-fatter, charge bleeding, and antiblooming nonlinearity are negligible or corrected elsewhere.
    Sec. 2.2 Eq. 2; Sec. 1 lists these effects as challenges but AstroWISP does not model them, so the demonstration implicitly assumes their contribution is small.
  • domain assumption The background is locally uniform within the annulus between the exclusion radius (6 px) and outer radius (13 px) after median and outlier rejection.
    Sec. 2.1, background determination paragraph; radii are tuned for PANOPTES cameras and could fail in crowded fields or strong gradients.
  • domain assumption Astrometric solutions derived from Gaia catalog matching provide source centers accurate to a small fraction of a pixel.
    Sec. 2.1 and companion paper Romero et al. 2025; the photometry model places the source center at the astrometric position, so subpixel astrometric errors propagate into flux errors.
  • domain assumption Pixel values are independent and their uncertainties follow Poisson statistics with known gain g_p.
    Sec. 2.2 error formula; flat-field corrections are assumed included in gain, and no covariance between pixels is accounted for.
  • ad hoc to paper The exact analytical expressions for integrals of arbitrary polynomials over circle-rectangle overlaps are valid and implemented to machine precision.
    Sec. 2, opening paragraph; no derivation or numerical verification is shown, so the reader must take the exactness claim on faith.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Tools for High Precision Photometry from Wide-Field Color Images." pith.science (2026). https://pith.science/paper/HBYFWJA3

@misc{pith2026250714289,
  author       = {Pith},
  title        = {Pith review of: Tools for High Precision Photometry from Wide-Field Color Images},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HBYFWJA3}},
  note         = {Machine review of arXiv:2507.14289}
}
read the original abstract

We present AstroWISP: a collection of image processing tools for source extraction, background determination, point spread function/pixel response function fitting, and aperture photometry. AstroWISP is particularly well-suited for working with detectors featuring a Bayer mask (an array of microfilters applied to each detector pixel to allow color photography), such as consumer DSLR cameras. Such detectors pose significant challenges for existing tools while offering a much cheaper alternative to specialized devices. As a result, consumer DSLR cameras with Bayer masks are often underutilized for precision photometry. \astrowisp{} addresses this limitation in an effort to democratize precision photometry and support broader community participation in research. We demonstrate that our tools produce high-precision photometry from such images, enabling the use of such devices for detecting exoplanet transits. We package our tools for all major operating systems to ensure accessibility for amateur astronomers.

Figures

Figures reproduced from arXiv: 2507.14289 by the authors.

Figure 1
Figure 1. Top: 88×88 pixel portion, centered on WASP-33, of a 5208×3476 pixel image taken with a DSLR camera. The pattern is due to the Bayer mask. Bottom: arrangement of pixels sensitive to different colors in super-pixels (highlighted area) in a DSLR detector. 2014)). Thirdly, the pixels typical of our target detectors can hold fewer electrons and their readout electronics have lower resolution. For example, professional ca… view at source ↗
Figure 2
Figure 2. WASP-33 b exoplanet transit, observed by PANOPTES (blue points and circles), TESS (red points), and theoretical lightcurve based on best known system pa￾rameters (green curve). The raw PANOPTES-DSLR mea￾surements, originating from the 4 color channels of four cam￾eras in Hawaii (Mauna Loa observatory) and California (Mt Wilson) are shown as blue points. The blue points are binned in time to create the blue circles a… view at source ↗
Figure 3
Figure 3. Posterior distribution of the transit tim￾ing, depth and duration of WASP-33 b using only the PANOPTES observations from [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Same as [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 4
Figure 4. Figure 4: The scatter (median absolute deviation from the median) of the individual channel lightcurves of PANOPTES observations of a 10◦ × 15◦ field centered on FU Orionis. From top to bottom: red channel, first green channel, second green channel, blue channel. 6 7 8 9 10 11 1…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 7 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter doi edition editor eprint howpublished institution journal key month number organization pages publisher school series title misctitle type volume year version url label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts ...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION format.url url empty "" new.block "" url * "" * if FUNCTION format.eprint eprint empty "" archivePrefix empty "" archivePrefix "arXiv" = new.block " " eprint * " " * new.block " " eprint * " " * if if if FUNCTION format.doi doi empty "" " " doi * " " * if FUNCTION format.pid doi empty eprint empty ur...

  3. [3]

    TX1bI1j4 5&hb -b

    thebibliography [1] 20pt to REFERENCES 6pt =0pt -12pt 10pt plus 3pt =0pt =0pt =1pt plus 1pt =0pt =0pt -12pt =13pt plus 1pt =20pt =13pt plus 1pt \@M =10000 =-1.0em =0pt =0pt 0pt =0pt =1.0em @enumiv\@empty 10000 10000 `\.\@m \@noitemerr \@latex@warning Empty `thebibliography' environment \@ifnextchar \@reference \@latexerr Missing key on reference command E...

    work page arXiv 2021

  4. [4]

    https://docs.github.com/en/actions

    2024, GitHub Actions . https://docs.github.com/en/actions

    work page 2024

  5. [5]

    2014, Journal of Instrumentation, 9, C03048, 10.1088/1748-0221/9/03/C03048

    Antilogus , P., Astier , P., Doherty , P., Guyonnet , A., & Regnault , N. 2014, Journal of Instrumentation, 9, C03048, 10.1088/1748-0221/9/03/C03048

    doi:10.1088/1748-0221/9/03/c03048 2014

  6. [6]

    P., Tollerud , E

    Astropy Collaboration , Robitaille , T. P., Tollerud , E. J., et al. 2013, , 558, A33, 10.1051/0004-6361/201322068

    doi:10.1051/0004-6361/201322068 2013

  7. [7]

    M., Sip o cz , B

    Astropy Collaboration , Price-Whelan , A. M., Sip o cz , B. M., et al. 2018, , 156, 123, 10.3847/1538-3881/aabc4f

    doi:10.3847/1538-3881/aabc4f 2018

  8. [8]

    M., Lim , P

    Astropy Collaboration , Price-Whelan , A. M., Lim , P. L., et al. 2022, , 935, 167, 10.3847/1538-4357/ac7c74

    doi:10.3847/1538-4357/ac7c74 2022

Show all 27 references
  1. [9]

    2024, cibuildwheel , 2.21.3

    cibuildwheel developers, & maintainers. 2024, cibuildwheel , 2.21.3. https://github.com/pypa/cibuildwheel

    2024

  2. [10]

    2011, , 197, 12, 10.1088/0067-0049/197/1/12

    Demory , B.-O., & Seager , S. 2011, , 197, 12, 10.1088/0067-0049/197/1/12

    2011 doi

  3. [11]

    M., Avsar , A., Peluso , D

    Esposito , T. M., Avsar , A., Peluso , D. O., et al. 2021, in Posters from the TESS Science Conference II (TSC2), 155, 10.5281/zenodo.5131297

    2021 doi

  4. [12]

    W., Lang , D., & Goodman , J

    Foreman-Mackey , D., Hogg , D. W., Lang , D., & Goodman , J. 2013, , 125, 306, 10.1086/670067

    2013 doi

  5. [13]

    Gaia Collaboration , Vallenari , A., Brown , A. G. A., et al. 2023, , 674, A1, 10.1051/0004-6361/202243940

    2023 doi

  6. [14]

    T., Guyon , O., Jovanovic , N., et al

    Gee , W. T., Guyon , O., Jovanovic , N., et al. 2021, Journal of Astronomical Telescopes, Instruments, and Systems, 7(4), 048001, 10.1117/1.JATIS.7.4.048001

    2021 doi

  7. [15]

    2012, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol

    Guyon , O., & Martinache , F. 2012, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 8444, Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, 10.1117/12.927201

    2012 doi

  8. [16]

    2014, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol

    Guyon , O., Walawender , J., Jovanovic , N., et al. 2014, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 9145, , 91453V, 10.1117/12.2057288

    2014 doi

  9. [17]

    2019, , 877, 109, 10.3847/1538-4357/ab1b1d

    Kawashima , Y., & Ikoma , M. 2019, , 877, 109, 10.3847/1538-4357/ab1b1d

    2019 doi

  10. [18]

    W., Mierle , K., Blanton , M., & Roweis , S

    Lang , D., Hogg , D. W., Mierle , K., Blanton , M., & Roweis , S. 2010, , 139, 1782, 10.1088/0004-6256/139/5/1782

    2010 doi

  11. [19]

    2020, Acta Astronautica, 166, 23, 10.1016/j.actaastro.2019.09.028

    Marchis , F., Malvache , A., Marfisi , L., Borot , A., & Arbouch , E. 2020, Acta Astronautica, 166, 23, 10.1016/j.actaastro.2019.09.028

    2020 doi

  12. [20]

    2012, , 421, 1825, 10.1111/j.1365-2966.2011.19813.x

    P \'a l , A. 2012, , 421, 1825, 10.1111/j.1365-2966.2011.19813.x

    2012

  13. [21]

    2015, MNRAS, 450, 3233, 10.1093/mnras/stv894

    Parviainen, H. 2015, MNRAS, 450, 3233, 10.1093/mnras/stv894

    2015 doi

  14. [22]

    Penev, K., Romero, A., & Gee, W. T. 2025, kpenev/AstroWISP: AstroWISP 1.0, 1.0, Zenodo, 10.5281/zenodo.14713146

    2025 doi

  15. [23]

    2018, Models of Star-Planet Magnetic Interaction , ed

    Strugarek , A. 2018, Models of Star-Planet Magnetic Interaction , ed. H. J. Deeg & J. A. Belmonte , 25, 10.1007/978-3-319-55333-7\_25

    2018 doi

  16. [24]

    R., et al

    Zellem , R., Biferno , A., Ciardi , D. R., et al. 2019, , 51, 416. 1903.07716

    2019 arXiv

  17. [25]

    T., Pearson , K

    Zellem , R. T., Pearson , K. A., Blaser , E., et al. 2020, , 132, 054401, 10.1088/1538-3873/ab7ee7

    2020 doi

  18. [26]

    \'A ., Penev , K., et al

    Zhang , M., Bakos , G. \'A ., Penev , K., et al. 2016, , 128, 035001, 10.1088/1538-3873/128/961/035001

    2016 doi

  19. [27]

    A., Kataria , T., et al

    Zhang , M., Knutson , H. A., Kataria , T., et al. 2018, , 155, 83, 10.3847/1538-3881/aaa458

    2018 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.