Pith. sign in

REVIEW 3 major objections 4 minor 24 references

Mitigating Polarization Leakage in Gas Pixel Detectors through Hybrid Machine Learning and Analytic Event Reconstruction

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper contends that a hybrid CNN-plus-moment-analysis reconstruction, validated on IXPE calibration beams, mitigates polarization leakage in extended-source data such as G21.5-0.9.

desk verdict Solid lab validation of a hybrid ML reconstruction for IXPE, but the real-data leakage-mitigation claim outruns the evidence. read the letter →

arxiv 2506.07828 v1 pith:ASQIXSBQ submitted 2025-06-09 astro-ph.IM

Nicoló Cibrario , Michela Negro , Raffaella Bonino , Nikita Moriakov , Luca Baldini , Niccoló Di Lalla , Alessandro Di Marco , Sergio Fabiani
show 13 more authors
Andrea Frassá Alessio Gorgi Fabio La Monaca Luca Latronico Simone Maldera Alberto Manfreda Fabio Muleri Nicola Omodei John Rankin Carmelo Sgró Stefano Silvestri Paolo Soffitta Stefano Tugliani
This is my paper
classification astro-ph.IM
keywords polarizationleakageX-raypolarimetryGasPixelDetectorIXPEconvolutionalneuralnetworkmomentanalysismodulationfactorspurious
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

The paper sets out to show that a hybrid reconstruction algorithm—a convolutional neural network that predicts where each photoelectron track began, feeding into the same analytic moment analysis used for IXPE data—works on real detector data and suppresses a systematic error called polarization leakage. Leakage is fake polarization that appears where an extended X-ray source has sharp intensity edges, because small errors in reconstructing the photon impact point bias the measured emission angles. The paper reports that the hybrid method matches the standard analysis on unpolarized calibration beams, improves the modulation factor by about 5 percent at 6.4 keV, and reduces the number of significantly polarized bins in a simulated unpolarized extended source from 67 to 4. Applied to the IXPE observation of the supernova remnant G21.5-0.9, it produces a polarization map whose outer regions differ markedly from the standard map, which the authors interpret as leakage being removed. If correct, this gives a practical way to reprocess IXPE observations of extended sources and obtain more reliable magnetic-field and emission-mechanism measurements.

What carries the argument

The load-bearing object is the hybrid track reconstruction: a convolutional neural network with hexagonal convolution layers is trained on simulated photoelectron track images to predict the photon impact point, and that predicted point is then substituted into the analytic moment-analysis pipeline, which weights pixel charges around the horseshoe region to assign each track an emission angle. The emission angles are converted to normalized Stokes parameters $q$ and $u$, and the polarization degree and angle follow from their magnitude and direction. The mechanism matters because polarization leakage is a consequence of impact-point misreconstruction: when the assumed absorption point is wrong, the inferred track geometry is biased and the Stokes parameters acquire a component tied to the local intensity gradient. Improving the impact point therefore attacks the root cause, while the spurious-modulation subtraction uses per-photon maps, measured in the lab, to remove detector-imposed Stokes offsets before the modulation factor is applied.

What would settle it

Take an unpolarized lab beam and place a mask with a sharp straight edge over the Gas Pixel Detector so the beam illuminates half the detector. If the hybrid method genuinely removes polarization leakage, the standard moment analysis will show the characteristic spurious polarization oriented perpendicular to the edge while the hybrid reconstruction will not, at equal exposure; a null result would refute the leakage-mitigation claim, while confirmation on real edge geometry would close the gap left by the uniform-beam calibration.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that the noise in the standard reconstruction is dominated by impact-point error, and that replacing the analytic impact point with a CNN-predicted one—while keeping the rest of the moment analysis untouched—removes most of that leakage without introducing new systematics. The evidence chain runs through calibration data: unpolarized beams reconstructed with the hybrid method show spurious-modulation maps essentially identical to the standard method's, and after subtracting those maps the residual intrinsic modulation is consistent across two detector orientations and across the three flight detector units. For polarized beams, the hybrid modulation factor is equal to or higher than the standard one at every energy, with the largest gain near 6.4 keV. The authors then apply the trained network to IXPE data of G21.5-0.9 and find that the outer regions of the polarization map, where leakage is expected to dominate, appear qualitatively different; combined with the simulated unpolarized extended-source test, they read this as direct evidence that the hybrid method mitigates leakage in a real observation.

Load-bearing premise

The central conclusion depends on the assumption that the differences between the standard and hybrid maps of G21.5-0.9 are caused by removal of polarization leakage rather than by distortion introduced by the algorithm; the lab validation does not exercise the sharp-edge geometry where leakage appears, and the only edge-geometry evidence comes from simulations of the same type used to train the network.

Editorial extensions

If this is right

  • Reprocessing IXPE extended-source observations with the hybrid method becomes feasible: the paper provides per-detector, per-year modulation-factor curves for all three flight units, so existing data can be re-analyzed without new observations.
  • Leakage-dominated outer regions of sources such as G21.5-0.9, which are precisely the regions used to probe magnetic-field structure, become usable for polarization studies instead of being masked or corrected ad hoc.
  • The hybrid method raises the modulation factor, so a given observation delivers a smaller statistical uncertainty on polarization degree and angle at energies where the gain is largest, notably around 6.4 keV.
  • The validation protocol—unpolarized beams for bias, polarized beams for response, rotated configurations to separate source and spurious signal—establishes a template for qualifying future machine-learning track reconstructions for space missions.
  • Since the network was trained on simulations but works on lab data without extra bias, the approach can be retrained or fine-tuned for detector pressure changes and applied to future GPD-based missions.

Reading between the lines

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

  • If impact-point accuracy is the true driver of leakage, the hybrid method should also reduce edge artifacts in total-intensity and spectral maps, not just polarization maps; this is a testable corollary the paper does not pursue.
  • The published comparison leaves open whether the hybrid map of G21.5-0.9 is strictly better than the standard map or merely different; an independent cross-check against a PSF-based leakage correction on the same observation would settle this.
  • A decisive lab test would use a sharp-edged mask over an unpolarized beam to create an artificial intensity edge across the GPD: the standard analysis should show the characteristic edge-parallel spurious polarization, and the hybrid method should not.
  • Because the simulation used to demonstrate leakage reduction comes from the same software family used to train the network, the quantitative factor (67 bins down to 4) could be optimistic; re-running the same test with an independent simulation or with real sharp-edge data would calibrate that number.
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 / 4 minor

Summary. The paper presents a hybrid reconstruction algorithm for the IXPE Gas Pixel Detector that combines a CNN (DenseNet-121 with hexagonal convolutions) with the analytic moment-analysis method. The CNN predicts the photon impact point; the hybrid method then computes event Stokes parameters. The authors report three main results: (i) a simulated unpolarized Crab-like extended source shows a reduction in significant spurious-polarization bins from 67 to 4 (Fig. 2); (ii) laboratory calibration data from unpolarized beams show no unexpected residual after spurious-modulation correction and show modulation factors consistent with simulation trends, with an improvement of about 5% at 6.4 keV (Figs. 6, 8); and (iii) application to the IXPE observation of G21.5-0.9 yields polarization maps that differ markedly from the standard analysis, which is interpreted as mitigation of polarization leakage (Fig. 9).

Significance. The paper's lab validation is a genuine advance: previous CNN-based reconstruction algorithms for IXPE showed large limitations when confronted with lab data, and here the hybrid method reproduces spurious-modulation maps, passes the 90-degree rotation consistency check, and follows simulation trends for the modulation factor. If the leakage-mitigation claim for real extended sources is confirmed, the method would be valuable for IXPE and future GPD missions. However, the load-bearing evidence for real-data leakage mitigation is currently an ixpesim simulation (same simulator family used for training) and an ungrounded comparison for G21.5-0.9; the uniform-beam lab data do not exercise the sharp-edge geometry that produces leakage. The stress-test concern therefore lands. With an independent edge-geometry validation, the paper could support its central claim.

major comments (3)
  1. [Section 2, Fig. 2] The simulated unpolarized Crab-like source test that yields the 67-to-4 reduction is generated with ixpesim v14.3.5, the same simulation framework used to produce the CNN training data (Section 2 and Cibrario et al. 2023). This is an in-family test and cannot independently certify behavior at sharp intensity edges in real detector data, especially given that the paper itself notes in Section 1 that prior CNN-based algorithms failed lab validation. Please add an independent edge-geometry check, for example a lab measurement with a sharp mask or knife edge illuminated by an unpolarized beam, or a simulation with a Monte Carlo code not used in training.
  2. [Section 4, Fig. 9] The G21.5-0.9 comparison has no ground truth. The 'marked difference in the structure of the polarization maps' is interpreted as leakage removal, but algorithm-induced distortion is a viable alternative: Appendix B states that the CNN tends to predict the impact point near the track barycenter below 3 keV, which changes the effective PSF and could alter the polarization map. Please provide a quantitative consistency check, such as analyzing a simulated G21.5-like source with known polarization through both pipelines, or comparing the hybrid map against the standard map with an independent leakage model (e.g., the sky-calibrated PSF approach of Dinsmore & Romani 2024). Without this, the abstract's word 'demonstrated' overstates what is shown.
  3. [Section 3.1, Table 2 and Fig. 6] The uniform-beam lab validation convincingly shows that the hybrid method introduces no large bias in flat illumination and that the spurious-modulation correction is consistent between the 90-degree rotated configurations. However, uniform illumination does not exercise the sharp-edge geometry in which polarization leakage appears. The conclusions in Section 5 should separate the validated uniform-beam properties from the unvalidated edge-geometry leakage claim, or the leakage claim should be supported with an edge-geometry test.
minor comments (4)
  1. [Caption of Fig. 1] 'Error bars are present but not visible' is not informative; please plot visible error bars or state the uncertainty level in the caption.
  2. [Header / metadata] The header says 'Draft version September 8, 2025' while the arXiv submission is dated June 9, 2025; please synchronize the version and date or add a changelog.
  3. [Section 4] Please specify the analysis details for G21.5-0.9 beyond the OBSID, including the energy range, event selection, and background treatment, so the comparison is reproducible.
  4. [Fig. 9 caption] The caption states that 2-sigma bins are marked in white and 3-sigma bins in red, while the text for Fig. 2 uses a 3-sigma threshold; please clarify the threshold used for the G21.5 maps and whether the same threshold appears in both figures.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the hybrid method's modulation and spurious-modulation performance are validated against independent laboratory beams, and the leakage-mitigation claim is applied to real IXPE data; the shared ixpesim simulator between training and the simulated leakage test is a validation limitation, not a circular reduction.

full rationale

The paper's derivation chain is not circular. The CNN is trained on ixpesim-generated photoelectron tracks to predict the photon impact point, and the hybrid reconstruction then feeds that impact point into the standard analytic moment analysis. The central quantitative claims for the method's reliability are tested on external experimental data: unpolarized and polarized calibration beams from the IXPE ground campaign (Section 3) provide independent measurements of spurious modulation and modulation factor, and the G21.5-0.9 observation (Section 4) is real IXPE data. The simulated extended-source test of Fig. 2 does use the same ixpesim software family that generated the CNN training set, so it is an in-distribution consistency check rather than an independent certification of edge-geometry behavior; however, the CNN was trained on impact-point labels, not on leakage maps or Stokes parameters, so the reduction in significant leakage bins is not a fitted parameter renamed as a prediction. The real-data G21.5 comparison has no ground truth and could in principle reflect algorithm-induced map distortion, but that is an evidence-strength limitation, not a circular argument. No quoted equation reduces to its own input, and no load-bearing premise is justified solely by a self-citation. The paper is self-contained against external benchmarks, so the circularity score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claims rest on calibration products fitted to lab data (modulation factor, spurious maps), on a trained CNN, and on the assumption that ixpesim simulations capture the detector behavior relevant to leakage. No new physical entities are introduced.

free parameters (3)
  • CNN weights (DenseNet-121 with hexagonal convolutions) = trained on ixpesim Monte Carlo tracks; weights not released
    The network's impact-point predictions are central to the hybrid method and are empirically fitted to simulated tracks.
  • Modulation factor curves per DU per year = reported in Fig. 8 for each DU, with three yearly versions per DU
    Measured from 100% polarized lab beams; used to convert measured Stokes parameters to polarization degree; not derived from first principles.
  • Spurious modulation maps q_SM and u_SM as functions of energy and position = 300x300 maps per energy beam and DU (e.g., Fig. 5), interpolated in energy
    Derived from unpolarized lab beams and used in Eq. 6 to correct event-by-event photon Stokes parameters.
assumptions (4)
  • domain assumption ixpesim Monte Carlo accurately models the GPD response (track images, charge, PSF) for training and leakage simulation.
    Used in Section 2 to train the CNN and to generate the unpolarized Crab mock observation; real-data validation is partial and limited to lab beam geometries.
  • domain assumption The Rankin et al. (2022) two-orientation method cleanly separates spurious modulation from intrinsic source polarization.
    Invoked in Section 3.1 to correct unpolarized lab beams; it assumes the spurious modulation phase is fixed in detector coordinates while the source polarization phase rotates by 90 degrees.
  • ad hoc to paper The simulated unpolarized Crab Nebula (CXO image plus IXPE PSF, no background) has zero true polarization, so any detected significant polarization is leakage.
    This is the basis of the quantitative leakage test in Section 2; if the simulation or PSF introduces residual polarization, the 67-to-4 bin reduction is contaminated.
  • ad hoc to paper Spurious modulation maps can be linearly interpolated in energy between the discrete lab beam energies.
    Footnote 1 in Section 3.1 states interpolation is used for events with energies not matching the beams; the validity of this interpolation is not demonstrated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mitigating Polarization Leakage in Gas Pixel Detectors through Hybrid Machine Learning and Analytic Event Reconstruction." pith.science (2026). https://pith.science/paper/ASQIXSBQ

@misc{pith2026250607828,
  author       = {Pith},
  title        = {Pith review of: Mitigating Polarization Leakage in Gas Pixel Detectors through Hybrid Machine Learning and Analytic Event Reconstruction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ASQIXSBQ}},
  note         = {Machine review of arXiv:2506.07828}
}
read the original abstract

Spatially resolved polarization measurements of extended X-ray sources are expanding our understanding of the emission mechanisms and magnetic field properties involved. Such measurements have been possible in the past few years thanks to the Imaging X-ray Polarimetry Explorer (IXPE). However, the analysis of extended sources suffers a systematic effect known as polarization leakage, which artificially affects the measured polarization signal. To address this issue, we built a hybrid reconstruction algorithm, which combines machine learning and analytic techniques to improve the reconstruction of photoelectron tracks in the Gas Pixel Detector and to significantly mitigate polarization leakage. This work presents the first application of this hybrid method to experimental data, including both calibration lab measurements and IXPE observational data. We confirmed the reliable performance of the hybrid method for both cases. Additionally, we demonstrated the algorithm's effectiveness in reducing the polarization leakage effect through the analysis of the IXPE observation of the supernova remnant G21.5-0.9. By enabling more reliable polarization measurements, this method can potentially yield deeper insights into the magnetic field structures, particle acceleration processes, and emission mechanisms at work within extended X-ray sources.

Figures

Figures reproduced from arXiv: 2506.07828 by the authors.

Figure 1
Figure 1. Upper panel: Modulation factor (𝜇) as a function of en￾ergy achieved with simulated data, for the standard moment analysis in black and for the hybrid method in red. Error bars are present but not visible. Lower panel: the ratio between the hybrid method 𝜇 and the standard one is reported as a gray dashed line. Our hybrid method also effectively reduced the systematic effect of the polarization leakage. In this rega… view at source ↗
Figure 2
Figure 2. Polarization leakage effect on a simulated unpolarized extended source. Upper panel: Polarization degree (color bar) and angle (red arrows) significantly (>3𝜎) detected by the standard mo￾ment analysis. Lower panel: Polarization degree and angle signifi￾cantly detected by the hybrid method. In both panels, the contours from the CXO observation are reported as white lines [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Emission angles distributions for three unpolarized en￾ergy beams, analyzed with DU1 at the 𝜖2 configuration (Upper panel: 2.04 keV; Middle panel: 2.98 keV; Lower panel: 5.89 keV). Results are reported for the hybrid method (red histogram) and for the stan￾dard moment analysis (black histogram). analysis. A clear residual polarization signal is detected in all the maps. The same maps are evaluated for all the source… view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Integrated spurious modulation for the central region (R < 3 mm) of each IXPE DU for each energy beam. Results are reported for both the standard moment analysis (upper panel) and the hybrid method (lower panel). position in the GPD where the photon is absorbed and to …
Figure 5
Figure 5. Figure 5: Normalized Stokes parameters as a function of the impact point position on the GPD plane achieved by the standard moment analysis (left panels) and by the hybrid method (right panels) for a 2.7 keV unpolarized energy beam, for a single detector unit onboard IXPE (DU1) …
Figure 6
Figure 6. Figure 6: Intrinsic source-induced modulation after the spurious modulation correction estimated with DU1 data. Results are re￾ported for each unpolarized energy beam, for both the hybrid method (𝜖1 configuration in red, 𝜖2 configuration in orange) and the stan￾dard moment analy…
Figure 7
Figure 7. Figure 7: Modulation factor (𝜇) achieved by the hybrid method (red) and standard moment analysis (black) for all DUs onboard of IXPE, and for each polarized energy beam and GPD configuration (FF or DFF). ods, particularly in the outer regions of the nebula, where the influence o…
Figure 9
Figure 9. Figure 9: Binned polarization maps from the IXPE observation of G21.5-0.9, generated using the standard moment analysis (upper panel) and the hybrid method (lower panel). The arrows orientation represents the polarization angle, and the color map corresponds to the polarization …
Figure 8
Figure 8. Figure 8: Modulation factor curves as a function of the energy achieved by the hybrid method (red) and standard moment analysis (black) for all DUs onboard of IXPE. Error bars are present but not visible. The ratios between the hybrid method 𝜇 and the standard one are reported f…
Figure 10
Figure 10. Figure 10: Intrinsic source-induced modulation after the spurious modulation correction achieved with DU2 and DU3 data. Results are reported for each unpolarized energy beam, for both the hybrid method (𝜖1 configuration in red, 𝜖2 configuration in orange) and the standard moment…
Figure 11
Figure 11. Figure 11: Left panel: normalized distribution of the hybrid method predicted emission angles for a 2.7 keV beam subsample of 𝜖1 configuration and DU1 data. Right panel: power spectrum of the emission angles distribution. In order to understand the origin of the high frequency p…
Figure 12
Figure 12. Figure 12: Upper left panel: Example of a photoelectron track as measured by the GPD. The yellow dot is the barycenter of the track, whereas the blue dot is the barycenter of the horseshoe region (marked by the dashed blue line). The dashed black line is the direction of the vec…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

24 extracted references · 4 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. [4]

    2021, Astroparticle Physics, 133, 102628, 10.1016/j.astropartphys.2021.102628

    Baldini , L., Barbanera , M., Bellazzini , R., et al. 2021, Astroparticle Physics, 133, 102628, 10.1016/j.astropartphys.2021.102628

    arXiv 2021

  4. [5]

    2022, SoftwareX, 19, 101194, https://doi.org/10.1016/j.softx.2022.101194

    Baldini, L., Bucciantini, N., Di Lalla , N., et al. 2022, SoftwareX, 19, 101194, https://doi.org/10.1016/j.softx.2022.101194

    arXiv 2022

  5. [6]

    2003, 4843, 383, 10.1117/12.459381

    Bellazzini , R., Angelini , F., Baldini , L., et al. 2003, 4843, 383, 10.1117/12.459381

    doi:10.1117/12.459381 2003

  6. [7]

    , Romani, R

    Bucciantini , N., Di Lalla, N. , Romani, R. W. R. , et al. 2023 a , A&A, 672, A66, 10.1051/0004-6361/202245744

    doi:10.1051/0004-6361/202245744 2023

  7. [8]

    2023 b , Nature Astronomy, 7, 602, 10.1038/s41550-023-01936-8

    Bucciantini , N., Ferrazzoli , R., Bachetti , M., et al. 2023 b , Nature Astronomy, 7, 602, 10.1038/s41550-023-01936-8

    doi:10.1038/s41550-023-01936-8 2023

  8. [9]

    , Moriakov, N

    Cibrario , N., Negro, M. , Moriakov, N. , et al. 2023, A&A, 674, A107, 10.1051/0004-6361/202346302

    doi:10.1051/0004-6361/202346302 2023

Show all 24 references
  1. [10]

    2019, PhD thesis, University of Pisa

    Di Lalla, N. 2019, PhD thesis, University of Pisa

    2019

  2. [11]

    2022 a , in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol

    Di Marco , A., Tennant , A., La Monaca , F., et al. 2022 a , in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 12181, Space Telescopes and Instrumentation 2022: Ultraviolet to Gamma Ray, ed. J.-W. A. den Herder , S. Nikzad , & K. Nakazawa , 1...

    2022 doi

  3. [12]

    2022 b , The Astronomical Journal, 164, 103, 10.3847/1538-3881/ac7719

    Di Marco , A., Fabiani, S., La Monaca , F., et al. 2022 b , The Astronomical Journal, 164, 103, 10.3847/1538-3881/ac7719

    2022 doi

  4. [13]

    T., & Romani, R

    Dinsmore, J. T., & Romani, R. W. 2024, The Astrophysical Journal, 962, 183, 10.3847/1538-4357/ad2065

    2024 doi

  5. [14]

    Huang, G., Liu, Z., Maaten, L. V. D., & Weinberger, K. Q. 2017, in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (Los Alamitos, CA, USA: IEEE Computer Society), 2261--2269, 10.1109/CVPR.2017.243

    2017 doi

  6. [15]

    2015, Astroparticle Physics, 68, 45, 10.1016/j.astropartphys.2015.02.007

    Kislat , F., Clark , B., Beilicke , M., & Krawczynski , H. 2015, Astroparticle Physics, 68, 45, 10.1016/j.astropartphys.2015.02.007

    2015 doi

  7. [16]

    2019, Nuclear Instruments and Methods in Physics Research A, 942, 162389, 10.1016/j.nima.2019.162389

    Kitaguchi , T., Black , K., Enoto , T., et al. 2019, Nuclear Instruments and Methods in Physics Research A, 942, 162389, 10.1016/j.nima.2019.162389

    2019

  8. [17]

    2022, Astroparticle Physics, 136, 102658, https://doi.org/10.1016/j.astropartphys.2021.102658

    Muleri, F., Piazzolla, R., Di Marco , A., et al. 2022, Astroparticle Physics, 136, 102658, https://doi.org/10.1016/j.astropartphys.2021.102658

    2022

  9. [18]

    L., Romani , R

    Peirson , A. L., Romani , R. W., Marshall , H. L., Steiner , J. F., & Baldini , L. 2021, Nuclear Instruments and Methods in Physics Research A, 986, 164740, 10.1016/j.nima.2020.164740

    2021

  10. [19]

    F., et al

    Rankin, J., Muleri, F., Tennant, A. F., et al. 2022, The Astronomical Journal, 163, 39, 10.3847/1538-3881/ac397f

    2022 doi

  11. [20]

    W., Wong, J., Di Lalla , N., et al

    Romani, R. W., Wong, J., Di Lalla , N., et al. 2023, The Astrophysical Journal, 957, 23, 10.3847/1538-4357/acfa02

    2023 doi

  12. [21]

    2021, The Astronomical Journal, 162, 208, 10.3847/1538-3881/ac19b0

    Soffitta, P., Baldini, L., Bellazzini, R., et al. 2021, The Astronomical Journal, 162, 208, 10.3847/1538-3881/ac19b0

    2021 doi

  13. [22]

    Steppa , C., & Holch , T. L. 2019, SoftwareX, 9, 193, 10.1016/j.softx.2019.02.010

    2019 doi

  14. [23]

    2000, Proc SPIE, 4012, 10.1117/12.391545

    Weisskopf, M., Tananbaum, H., Speybroeck, L., & O'Dell, S. 2000, Proc SPIE, 4012, 10.1117/12.391545

    2000 doi

  15. [24]

    C., Soffitta, P., Baldini, L., et al

    Weisskopf, M. C., Soffitta, P., Baldini, L., et al. 2022, Journal of Astronomical Telescopes, Instruments, and Systems, 8, 026002, 10.1117/1.JATIS.8.2.026002

    2022 doi

  16. [25]

    2022, , 612, 658, 10.1038/s41586-022-05476-5

    Xie , F., Di Marco , A., La Monaca , F., et al. 2022, , 612, 658, 10.1038/s41586-022-05476-5

    2022 doi

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.