Pith. sign in

REVIEW 3 major objections 5 minor 69 references

PyTICS: An Iterative Method for Photometric Lightcurve Intercalibration using Comparison Stars

T0 review · 3 major / 5 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 An iterative algorithm intercalibrates multi-telescope photometric lightcurves using hundreds of comparison stars alone, without modelling the target's variability, and adds a colour correction for blue AGNs.

desk verdict A solid, reproducible intercalibration method that genuinely improves uncertainty handling; the color-correction extrapolation is a real but openly acknowledged limitation, not a fatal flaw. read the letter →

arxiv 2505.23328 v2 pith:V5BMZOJU submitted 2025-05-29 astro-ph.IM astro-ph.GA

Roberta Vieliute , Juan V. Hernández Santisteban , Keith Horne , Hannah Cornfield This is my paper
classification astro-ph.IMastro-ph.GA
keywords photometricintercalibrationensemblephotometryactivegalacticnucleireverberationmappinglightcurvecalibrationextravariancenoisemodelcomparisonstarscolour-dependentsystematics
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

PyTICS is a new iterative algorithm for intercalibrating photometric time series taken with many telescopes: it uses the hundreds of comparison stars present on the same images to estimate telescope-specific and epoch-specific magnitude corrections, then applies those corrections to the AGN at the end. The paper's central claim is that no assumption about the shape of the AGN variability is needed, because the target is never modelled; instead, scatter in the comparison-star residuals drives a multi-component noise model that inflates nominal uncertainties where epochs, telescopes, or stars behave badly. Demonstrated on ten 1-m telescopes of a global robotic observatory network observing the Seyfert galaxy NGC3783, the method yields lightcurves with realistic error bars and identifies outlier epochs and telescopes more cleanly than the comparison tool PyROA. The paper also shows that residual trends with star colour matter for AGNs bluer than their field stars, and introduces a first-order colour correction that removes most of the remaining telescope splitting for Fairall9 and three other blue AGNs. A sympathetic reader would care because reverberation-mapping delays and their uncertainties depend directly on clean, correctly weighted multi-telescope lightcurves.

What carries the argument

The load-bearing object is the additive magnitude model $m(\star,\mathrm{Tel},\mathrm{Ep}) = \bar{m}_\star + \Delta m_{\mathrm{Tel}} + \Delta m_{\mathrm{Ep}}$ combined with a four-component noise model $\sigma^2(\star,\mathrm{Tel},\mathrm{Ep}) = \sigma_i^2 + \sigma_\star^2 + \sigma_{\mathrm{Tel}}^2 + \sigma_{\mathrm{Ep}}^2$, where $\sigma_i^2$ is the nominal variance. Each set of parameters is estimated by maximising the Gaussian likelihood through the Badness-of-Fit statistic of Eq. (4), with a 'goodness' weight $g_i = \sigma_x^2/(\sigma_x^2+(\sigma_i')^2)$ that down-weights data whose extra variance is less informative; the parameter updates iterate until changes fall below a small fraction of their uncertainties. The same machinery identifies bad epochs (large $\sigma_{\mathrm{Ep}}$), bad telescopes (large $\sigma_{\mathrm{Tel}}$), and intrinsically variable stars (large $\sigma_\star$), so that corrections transferred to the AGN come with an empirically grounded error model.

What would settle it

Observe an AGN bluer than all its field stars on a night when the same field is also observed by a well-calibrated reference, or generate synthetic images with known non-linear colour terms; if the PyTICS colour correction leaves telescope splitting at the level of the colour slopes, or recovers an injected offset incorrectly, the linear extrapolation of star residuals to the target is falsified.

Watch

Extended reading notes

Core claim

The paper introduces an iterative optimal scaling algorithm that fits the model $m(\star,\mathrm{Tel},\mathrm{Ep}) = \bar{m}_\star + \Delta m_{\mathrm{Tel}} + \Delta m_{\mathrm{Ep}}$ to uncalibrated instrumental magnitudes of many comparison stars, with a noise model $\sigma^2 = \sigma_i^2 + \sigma_\star^2 + \sigma_{\mathrm{Tel}}^2 + \sigma_{\mathrm{Ep}}^2$. It solves for each parameter set by maximising a Gaussian likelihood, iterating over epoch corrections, telescope corrections, and star magnitudes until convergence, then transfers the summed correction parameters to the AGN lightcurve. The claim is that this makes calibration independent of the AGN's stochastic variability, avoids interpolation and smoothing biases, and produces defensible uncertainties: normalised residuals of the calibrated comparison stars form a Gaussian centred at zero, and the noise model flags problematic epochs, telescopes, and stars automatically. On NGC3783 data the PyTICS lightcurve matches the variability pattern of the PyROA-calibrated curve, but with epoch-resolved extra variance that captures the periods when one telescope had focus issues without inflating all that telescope's points. For blue AGNs like Fairall9, whose colour lies outside the comparison-star range, the paper adds a first-order colour correction: a Bayesian linear fit to mean residual versus star colour, extrapolated to the AGN colour and applied per telescope, which reduces the remaining splitting for Fairall9, PG1119+120, 3C273, and Mrk1044.

Load-bearing premise

The whole calibration assumes that telescope and epoch offsets are the same for the AGN and for the comparison stars; for AGNs bluer than all their field stars it also assumes that the residual-versus-colour trend can be safely extrapolated linearly past the last star.

Editorial extensions

If this is right

  • Reverberation-mapping lightcurves from heterogeneous telescope networks can be intercalibrated without assuming a damped random walk or a running optimal average, so high-frequency variability and power-spectral-density estimates are not smoothed away.
  • Outlier epochs and telescopes are automatically down-weighted by epoch- and telescope-specific extra variances, so a single problematic telescope need not inflate the uncertainties of its good data or distort lag measurements.
  • The comparison-star noise model yields a wider, unimodal AGN uncertainty distribution than PyROA's clipped, bimodal extra variance, which the paper argues better protects cross-correlation and lag-fitting results.
  • For targets bluer than all field stars, extrapolating the residual-versus-colour slope to the AGN colour removes most telescope splitting in Fairall9, PG1119+120, 3C273, and Mrk1044, with a global solution available across the network.
  • The method costs far less compute than MCMC-based intercalibration, since it solves the likelihood equations iteratively rather than sampling a full joint posterior.

Reading between the lines

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

  • If the residual-versus-colour relation is not linear across the full stellar colour range, extrapolating it to an AGN bluer than every comparison star will be biased; this could be tested with fields whose stars bracket the AGN colour or with synthetic images that inject a known non-linear colour term.
  • Because the AGN is never modelled, the same pipeline should transfer to other variable sources, such as supernovae, variable stars, or transients, whenever enough comparison stars are present in the field.
  • A natural next step implied by the paper's caveats is to fold the colour term into the iterative loop by solving two filters jointly, yielding per-epoch, per-telescope colour-dependent corrections and a colour-dependent extra variance.
  • The noise model assumes no covariance among its components; simulated epochs with correlated telescope drift would reveal where that assumption limits the accuracy of the quoted error bars.
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 presents PyTICS, an iterative maximum-likelihood algorithm for intercalibrating multi-telescope photometric light curves of AGNs using 10s to 100s of comparison stars in the same field. The algorithm estimates telescope-specific and epoch-specific magnitude offsets, plus a multi-component noise model containing star-, telescope-, and epoch-specific extra variances. Calibration parameters are estimated from the comparison stars and then applied to the AGN light curve, avoiding any assumption about the AGN variability shape or any interpolation of the AGN data. The method is demonstrated on LCO 1-m observations of NGC 3783 and compared with PyROA, and the paper also investigates residual colour-dependent trends, proposing a first-order colour correction for unusually blue AGNs such as Fairall 9.

Significance. If the method performs as claimed, it addresses a genuine and practical problem in intensive reverberation-mapping campaigns, where combining multi-site photometry requires careful intercalibration. The approach of deriving corrections from an ensemble of comparison stars is conceptually appealing and may improve robustness against outliers and underestimated uncertainties. Strong points include the public availability of the Python package, the explicit noise model with extra variance parameters, and the demonstration that the algorithm identifies problematic epochs and telescopes. The colour-correction extension for blue AGNs, while not fully validated, highlights an interesting systematic effect. The claim that the method requires no assumptions about the AGN variability is correct for the main procedure, and the computational cost appears favorable compared with MCMC-based approaches.

major comments (3)
  1. [Section 4.1, Fig. 10, Table 1] The colour-correction step for blue AGNs relies on extrapolating a linear fit of mean residual versus star colour to AGN colours outside the comparison-star range. For Fairall 9 in the B band, the AGN has B-V = 0.18 while the 92 field stars span 0.30-1.45, so the correction is entirely extrapolated. The paper does not quantify the residual telescope splitting before and after the colour correction, and it does not test the linearity assumption or the choice of u-g colour index on simulated data with known colour-dependent offsets. This is load-bearing for the abstract claim that the algorithm 'can in principle be applied to any astronomical object'; a small error in the extrapolated slope could leave telescope-specific offsets that mimic short-timescale AGN variability. Please provide either a synthetic test with injected colour-dependent offsets, or a quantitative evaluation of the residual scatter before and after correction, and discuss the sensitivity to the colour index choice.
  2. [Section 3.2, Figs. 5-6] The comparison with PyROA is performed only on real data with unknown ground truth. The paper claims that PyTICS 'can more accurately quantify the uncertainties', but this is not demonstrated by the presented evidence. A simulation with known input uncertainties and known telescope/epoch offsets would allow a direct test of whether the noise model returns unbiased estimates of the star-, telescope-, and epoch-specific extra variances and whether the AGN error bars are correctly calibrated. Without such a test, the claim of superior uncertainty quantification rests on distributional arguments that are partly self-referential. I recommend adding a simulation-based validation, at least for the noise model component.
  3. [Section 2.1, Eq. (3)-(8), Fig. 3] The validation of the noise model via normalized residuals is partly circular: the extra variance parameters are fit by maximizing the likelihood on the same data that are then used to check that the residuals are Gaussian. The Gaussianity of the normalized residuals is a necessary but not sufficient check. A more convincing validation would be to hold out a subset of stars or epochs during the fit and then evaluate the predictive performance on the held-out data, or to compare the estimated star-specific variances against the RMS versus magnitude scatter in a more quantitative way.
minor comments (5)
  1. [Section 2.2] The convergence criterion 'a small fraction (e.g. 10^-5) of the corresponding parameter uncertainties' is stated in words; it would be helpful to give the explicit equation used in the code.
  2. [Section 3.1, Fig. 3] The histogram of normalized residuals is only visual; adding a Kolmogorov-Smirnov test or a reported chi-squared value would make the Gaussianity claim more quantitative.
  3. [Section 3.1] The choice of 100 brightest comparison stars is described as 'arbitrary'; it would strengthen the paper to show how the final intercalibration parameters and uncertainties depend on this number and on the completeness threshold.
  4. [Section 4.1] The paper acknowledges that the colour index choice is 'somewhat arbitrary' and that u-g worked better than other indices. Please state explicitly whether this choice was made after inspecting the data, and note the risk of selection bias when reporting the successful corrections.
  5. [Appendix A] The global colour-correction solution from 32 AGN fields shows consistent deviations for telescopes 1m004 and 1m010. A brief comment on the possible origin of these deviations would be useful.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: star-derived calibration is applied to the AGN without using AGN variability, and the colour correction is an extrapolation, not a refit.

full rationale

The central derivation is self-contained. Equation (2) models comparison-star magnitudes as star-specific means plus telescope- and epoch-specific additive corrections, and all correction parameters and extra-variance terms are estimated from comparison-star residuals via the maximum-likelihood equations (4)-(8). The AGN lightcurve is deliberately excluded from the fit (footnote 7), so the calibrated AGN variability is not forced by construction. The Fairall 9 colour term is likewise not circular: the linear residual-versus-colour slopes are fit only to field-star residuals (Fig. 8) and then extrapolated to the AGN colour. This is an extrapolation with admitted assumptions (linearity, and a colour index described as 'somewhat arbitrary'), which are limitations rather than self-referential reductions. The normalized-residual histogram (Fig. 3 right) is an in-sample goodness-of-fit diagnostic of the same star residuals used to estimate the noise model; because it is presented as a quality check rather than an independent prediction, it does not constitute circularity. Self-citations such as Hernandez Santisteban et al. (2020) are used for data-reduction details and examples of telescope splitting, not as load-bearing derivations. No fitted parameter is renamed as a prediction, and no central claim reduces to its own inputs.

Assumptions & free parameters 11 free parameters · 8 assumptions · 0 invented entities

The central model uses two classes of fitted parameters: magnitude shifts (star, telescope, epoch) and extra variances (star, telescope, epoch). The extra variance parameters are the key innovation and are not independently benchmarked. Several hand-set decisions enter the demonstrated pipeline: 100 stars, a 50 percent completeness threshold, a convergence threshold, the u-g colour index, and linear colour extrapolation. No new physical entities are introduced.

free parameters (11)
  • Per-star mean magnitude mbar_star = estimated per star
    Equation (2); the mean level of each comparison star, fitted by maximum likelihood in step (iii).
  • Telescope offset Delta_m_Tel = per-telescope values, e.g., Table 1 for B band
    Equation (2); correction for each telescope, fitted in step (ii) and normalized to zero mean.
  • Epoch offset Delta_m_Ep = per-epoch values
    Equation (2); per-epoch atmospheric and instrumental correction, fitted in step (i).
  • Star extra variance sigma_star^2 = per star
    Equation (3); down-weights intrinsically variable stars; fitted in step (iii).
  • Telescope extra variance sigma_Tel^2 = per telescope, e.g., 0.0074 for 1m004 in Table 1
    Equation (3); absorbs telescope-dependent scatter.
  • Epoch extra variance sigma_Ep^2 = per epoch
    Equation (3); absorbs epoch-specific scatter, e.g., the focus issues of 1m009.
  • Number of comparison stars = 100 brightest
    Chosen by hand in Section 3.1; using more does not significantly affect the final parameter estimates.
  • Completeness threshold for comparison stars = 50 percent of maximum datapoints
    Chosen in Section 3.1 to remove poorly sampled stars from the ensemble.
  • Convergence threshold = 1e-5 of parameter uncertainties
    Stopping criterion for the iterative loop, Section 2.2.
  • Colour-residual slope per telescope = Table 1 lists B-band slopes, e.g., -0.0454 +/- 0.0042 for 1m004
    Fitted with linmix in Section 4.1 and extrapolated to the AGN colour to form Delta_m_Col,Tel.
  • Colour index used for colour correction = u-g
    Chosen in Section 4.1 after trying several indices; the paper notes the choice is arbitrary and may not work for all filters.
assumptions (8)
  • domain assumption Comparison stars are constant sources: after removing mean, telescope, and epoch offsets, residual scatter is noise.
    Core to Eq. (2); variable stars are handled by down-weighting via sigma_star^2, so intrinsic variability is allowed but treated as noise. Section 2.1.
  • domain assumption Systematic offsets are additive in magnitude and identical for the AGN and the comparison stars.
    Equation (2) and transfer of Delta_m_Tel and Delta_m_Ep to the AGN in Section 2.2; colour-dependent residuals in Section 4 show this is only approximately true.
  • domain assumption Noise components add in quadrature with no covariance.
    Equation (3) and the statement 'This model assumes no covariance among the noise model parameters.'
  • standard math Magnitude residuals are Gaussian distributed.
    Justifies the -2 ln L likelihood in Eq. (4).
  • standard math The zero-mean normalization on Delta_m_Tel and Delta_m_Ep removes the degeneracy in Eq. (2).
    Section 2.1; otherwise the model has an infinite family of solutions.
  • ad hoc to paper The iterative coordinate-wise updates converge to the maximum-likelihood solution.
    Section 2.2 states convergence is reached when parameters change by less than 1e-5 of their uncertainties, but no proof or synthetic test is given.
  • ad hoc to paper Residuals versus star colour are linear and can be extrapolated to AGN colours outside the comparison-star range.
    Section 4.1 and Fig. 8; used to compute Delta_m_Col,Tel for blue AGNs like Fairall 9.
  • ad hoc to paper The u-g colour index is a suitable basis for the first-order colour correction.
    Section 4.1: 'The choice of colour index is thus somewhat arbitrary and may not work as well if only certain filters are available.'

how reviews work

0 comments
Cite this review

Pith. "Pith review of PyTICS: An Iterative Method for Photometric Lightcurve Intercalibration using Comparison Stars." pith.science (2026). https://pith.science/paper/V5BMZOJU

@misc{pith2026250523328,
  author       = {Pith},
  title        = {Pith review of: PyTICS: An Iterative Method for Photometric Lightcurve Intercalibration using Comparison Stars},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/V5BMZOJU}},
  note         = {Machine review of arXiv:2505.23328}
}
read the original abstract

Intensive reverberation mapping monitoring programs combine ground-based photometric observations from different telescopes, requiring intercalibration of lightcurves to reduce systematic instrumental differences. We present a new iterative algorithm to calibrate photometric time-series data of active galactic nuclei (AGN) using 100s of comparison stars on the same images, building upon the established method of ensemble photometry. The algorithm determines telescope-specific and epoch-specific correction parameters, and simultaneously computes a multi-component noise model to account for underestimated uncertainties based on the scatter in the comparison star data, effectively identifying problematic epochs, telescopes, and stars. No assumptions need to be made about the AGN variability shape, and the algorithm can in principle be applied to any astronomical object. We demonstrate our method on lightcurves taken with ten 1-m telescopes from the Las Cumbres Observatory (LCO) robotic telescope network. Comparing our results to other intercalibration tools, we find that the algorithm can more accurately quantify the uncertainties in the data. We describe additional corrections that can be made for particularly bluer AGNs like Fairall 9, arising due to systematic effects dependent on star colour.

Figures

Figures reproduced from arXiv: 2505.23328 by the authors.

Figure 1
Figure 1. RMS as a function of 𝑚¯ ★ for 445 field stars around NGC 3783 in the LCO gp band, where 𝑚¯ ★ is the mean instrumental star magnitude derived from our algorithm. The hundred brightest stars (green) are selected for the intercalibration. The RMS of each star shown here is the mean of the nominal uncertainties (𝜎¯ 2 𝑖 = 1 𝑁Ep Í 𝑖 𝜎2 𝑖 ) added in quadrature with the star￾specific extra variance parameter derived from ou… view at source ↗
Figure 2
Figure 2. Schematic of the iterative intercalibration algorithm PyTICS, demonstrated on a section of NGC 3783 data for the LCO up band. Datapoints are coloured by telescope. Top: Datapoints are grouped by epoch (coloured group at each MJD) after subtracting the mean telescope magnitude for each star. The optimal values of 𝛿𝑚Ep and 𝜎2 Ep are computed given each epoch residual cluster and the outlined noise model. Middle: Datap… view at source ↗
Figure 3
Figure 3. Top: Uncalibrated instrumental magnitude lightcurve for a non-variable star in the field of NGC 3783, for the LCO ip band. The datapoints are coloured according to the telescopes in the LCO network, and show significant offsets as well as errorbars not representative of the scatter. Bottom: Intercalibrated lightcurve using PyTICS. The correction parameters derived from the algorithm bring the data to a mean magnitud… view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Top: Uncalibrated instrumental magnitude lightcurve for NGC 3783 in the LCO ip band, with datapoints coloured according to the telescope. Middle: Intercalibrated lightcurve using PyTICS. The variability pattern is more representative of AGN variations. The extreme outl…
Figure 5
Figure 5. Figure 5: A section of the NGC 3783 lightcurve for the LCO ip band, inter￾calibrated using PyROA (top) and PyTICS (bottom). Datapoints are coloured according to the telescope. The large pink outliers are from telescope 1m009, which experienced focus issues during this period [P…
Figure 6
Figure 6. Figure 6: Distribution of the total noise model in magnitudes for NGC 3783 in the LCO ip band, derived from PyTICS (green) and PyROA (red), together with the cumulative distribution functions (CDF). lag measurements but the lag uncertainties as well, which can in turn result in …
Figure 7
Figure 7. Figure 7: Normalised residual heatmap of the comparison stars used for the calibration of NGC 3783 in the LCO B band. Epochs are sorted by telescope and stars are sorted by 𝐵 − 𝑉 colour, showing clear residual trends with different slopes for each telescope. have slopes that are…
Figure 8
Figure 8. Figure 8: Top: The mean (non-normalised) residuals as a function of the 𝐵 − 𝑉 colour index for the field stars of Fairall 9 (red), NGC 3783 (blue), and NGC 7469 (purple), in the LCO 𝐵 band. Vertical dashed lines indicate the mean AGN colour, with horizontal errorbars showing the…
Figure 9
Figure 9. Figure 9: Same as [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 10
Figure 10. Figure 10: Top: A section of the Fairall 9 lightcurve in the LCO 𝐵 band, coloured by telescope, showing strong splitting even after using PyTICS. A first-order colour correction can be derived from fitting the mean (non￾normalised) residuals as a function of star colour and extr…
Figure 11
Figure 11. Figure 11: Same as [PITH_FULL_IMAGE:figures/full_fig_p010_11.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

69 extracted references · 35 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION fin.entry write newline FUNCTION new.block output.state before.all = 'skip after.block 'output.state := if FUNCTION new.sentence output.state after.block = 'skip output.state before.all = 'skip after.sentence 'output.state := if if FUNCTION not #0 #1 if FUNCTION and 'skip pop #0 if FUNCTION or pop #1...

  2. [2]

    D., McKee C

    Blandford R. D., McKee C. F., 1982, @doi [ApJ] 10.1086/159843 , 255, 419

    doi:10.1086/159843 1982

  3. [3]

    M., et al., 2013, @doi [PASP] 10.1086/673168 , 125, 1031

    Brown T. M., et al., 2013, @doi [PASP] 10.1086/673168 , 125, 1031

    doi:10.1086/673168 2013

  4. [4]

    M., Chiang C.-Y., McHardy I., Edelson R., Goad M

    Cackett E. M., Chiang C.-Y., McHardy I., Edelson R., Goad M. R., Horne K., Korista K. T., 2018, @doi [ApJ] 10.3847/1538-4357/aab4f7 , 857, 53

    doi:10.3847/1538-4357/aab4f7 2018

  5. [5]

    M., Bentz M

    Cackett E. M., Bentz M. C., Kara E., 2021, @doi [iScience] 10.1016/j.isci.2021.102557 , 24, 102557

    arXiv 2021

  6. [6]

    J., et al., 1998, @doi [ApJ] 10.1086/305720 , 500, 162

    Collier S. J., et al., 1998, @doi [ApJ] 10.1086/305720 , 500, 162

    doi:10.1086/305720 1998

  7. [7]

    R., Horne K., Hernández Santisteban J

    Donnan F. R., Horne K., Hernández Santisteban J. V., 2021, @doi [MNRAS] 10.1093/mnras/stab2832 , 508, 5449

    doi:10.1093/mnras/stab2832 2021

  8. [8]

    R., et al., 2023, @doi [MNRAS] 10.1093/mnras/stad1409 , 523, 545

    Donnan F. R., et al., 2023, @doi [MNRAS] 10.1093/mnras/stad1409 , 523, 545

    doi:10.1093/mnras/stad1409 2023

Show all 69 references
  1. [9]

    Edelson R., et al., 2015, @doi [ApJ] 10.1088/0004-637X/806/1/129 , 806, 129

    2015 doi

  2. [10]

    M., et al., 2016, @doi [ApJ] 10.3847/0004-637X/821/1/56 , 821, 56

    Fausnaugh M. M., et al., 2016, @doi [ApJ] 10.3847/0004-637X/821/1/56 , 821, 56

    2016 doi

  3. [11]

    M., Peterson B

    Fausnaugh M. M., Peterson B. M., Starkey D. A., Horne K., the AGN STORM Collaboration 2017, @doi [Front. Astron. Space Sci.] 10.3389/fspas.2017.00055 , 4, 55

    2017

  4. [12]

    Fernández Fernández J., Chou D.-Y., Pan Y.-C., Wang L.-H., 2012, @doi [PASP] 10.1086/665926 , 124, 507

    2012 doi

  5. [13]

    Gardner E., Done C., 2017, @doi [MNRAS] 10.1093/mnras/stx946 , 470, 3591

    2017 doi

  6. [14]

    Hagen S., Done C., Edelson R., 2024, @doi [MNRAS] 10.1093/mnras/stae1177 , 530, 4850

    2024 doi

  7. [15]

    V., et al., 2020, @doi [MNRAS] 10.1093/mnras/staa2365 , 498, 5399

    Hernández Santisteban J. V., et al., 2020, @doi [MNRAS] 10.1093/mnras/staa2365 , 498, 5399

    2020 doi

  8. [16]

    K., 1992, @doi [PASP] 10.1086/133015 , 104, 435

    Honeycutt R. K., 1992, @doi [PASP] 10.1086/133015 , 104, 435

    1992 doi

  9. [17]

    Hunter, J. D. 2007, Computing In Science & Engineering , 9, 90

    2007

  10. [18]

    S., Papadakis I

    Kammoun E. S., Papadakis I. E., Dovčiak M., 2021, @doi [MNRAS] 10.1093/mnras/stab725 , 503, 4163

    2021 doi

  11. [19]

    Kara E., et al., 2021, @doi [ApJ] 10.3847/1538-4357/ac2159 , 922, 151

    2021 doi

  12. [20]

    P., Vogeley M

    Kasliwal V. P., Vogeley M. S., Richards G. T., 2015, @doi [MNRAS] 10.1093/mnras/stv1230 , 451, 4328

    2015 doi

  13. [21]

    C., 2007, @doi [ ] 10.1086/519947 , https://ui.adsabs.harvard.edu/abs/2007ApJ...665.1489K 665, 1489

    Kelly B. C., 2007, @doi [ ] 10.1086/519947 , https://ui.adsabs.harvard.edu/abs/2007ApJ...665.1489K 665, 1489

    2007 doi

  14. [22]

    C., Bechtold J., Siemiginowska A., 2009, @doi [ApJ] 10.1088/0004-637X/698/1/895 , 698, 895

    Kelly B. C., Bechtold J., Siemiginowska A., 2009, @doi [ApJ] 10.1088/0004-637X/698/1/895 , 698, 895

    2009 doi

  15. [23]

    T., Goad M

    Korista K. T., Goad M. R., 2019, @doi [MNRAS] 10.1093/mnras/stz2330 , 489, 5284

    2019 doi

  16. [24]

    H., Horne K., Kallman T

    Krolik J. H., Horne K., Kallman T. R., Malkan M. A., Edelson R. A., Kriss G. A., 1991, @doi [ApJ] 10.1086/169918 , 371, 541

    1991 doi

  17. [25]

    Li Y.-R., Wang J.-M., Hu C., Du P., Bai J.-M., 2014, @doi [ApJ] 10.1088/2041-8205/786/1/L6 , 786, L6

    2014 doi

  18. [26]

    Netzer H., 2020, @doi [MNRAS] 10.1093/mnras/staa767 , 494, 1611

    2020 doi

  19. [27]

    M., 1993, @doi [PASP] 10.1086/133140 , 105, 247

    Peterson B. M., 1993, @doi [PASP] 10.1086/133140 , 105, 247

    1993 doi

  20. [28]

    G., Doroshenko V

    Sergeev S. G., Doroshenko V. T., Golubinskiy Y. V., Merkulova N. I., Sergeeva E. A., 2005, @doi [ApJ] 10.1086/427820 , 622, 129

    2005 doi

  21. [29]

    I., Sunyaev R

    Shakura N. I., Sunyaev R. A., 1973, Astronomy and Astrophysics, 24, 337

    1973

  22. [30]

    Shen Y., et al., 2024, @doi [ApJS] 10.3847/1538-4365/ad3936 , 272, 26

    2024 doi

  23. [31]

    Simm T., Salvato M., Saglia R., Ponti G., Lanzuisi G., Trakhtenbrot B., Nandra K., Bender R., 2016, @doi [A&A] 10.1051/0004-6361/201527353 , 585, A129

    2016 doi

  24. [32]

    A., Huang J., Horne K., Lin D

    Starkey D. A., Huang J., Horne K., Lin D. N. C., 2022, @doi [MNRAS] 10.1093/mnras/stac3579 , 519, 2754

    2022 doi

  25. [33]

    Sun M., et al., 2020, @doi [ApJ] 10.3847/1538-4357/ab789e , 891, 178

    2020 doi

  26. [34]

    Tamuz O., Mazeh T., Zucker S., 2005, @doi [MNRAS] 10.1111/j.1365-2966.2004.08585.x , 356, 1466

    2005

  27. [35]

    The pandas development Team 2024, pandas-dev/pandas: Pandas , @doi 10.5281/zenodo.3509134

    2024 doi

  28. [36]

    Theureau G., Bottinelli L., Coudreau-Durand N., Gouguenheim L., Hallet N., Loulergue M., Paturel G., Teerikorpi P., 1998, @doi [ ] 10.1051/aas:1998416 , https://ui.adsabs.harvard.edu/abs/1998A&AS..130..333T 130, 333

    1998 doi

  29. [37]

    M., Padovani P., 1995, @doi [PASP] 10.1086/133630 , 107, 803

    Urry C. M., Padovani P., 1995, @doi [PASP] 10.1086/133630 , 107, 803

    1995 doi

  30. [38]

    and McKee, C., BM82 1982 , ApJ, 255, 419-439

    Blandford, R. and McKee, C., BM82 1982 , ApJ, 255, 419-439

    1982

  31. [39]

    Brown T. M. et al., Brown13 2013 , PASP, 125, 1031

    2013

  32. [40]

    Cackett, E., Bentz, M., and Kara, E., Cackett21 2021 , iScience, 24, 6

    2021

  33. [41]

    Cackett18 2018 , ApJ, 857, 53

    Cackett, E., Chiang, C., McHardy, I., et al. Cackett18 2018 , ApJ, 857, 53

    2018

  34. [42]

    Collier98 1998 , ApJ, 500, 162-172

    Collier, S., Horne, K., Kaspi, S., et al. Collier98 1998 , ApJ, 500, 162-172

    1998

  35. [43]

    Donnan23 2023 , MNRAS, 523, 545–567

    Donnan, F., Hernández Santisteban, J., Horne, K., et al. Donnan23 2023 , MNRAS, 523, 545–567

    2023

  36. [44]

    Donnan, F., Horne, K., and Hernández Santisteban, J., Donnan21 2021 , MNRAS, 508, 5449–5467

    2021

  37. [45]

    Edelson15 2015 , ApJ, 806, 129

    Edelson, R., Gelbord, J., Horne, K., et al. Edelson15 2015 , ApJ, 806, 129

    2015

  38. [46]

    Fausnaugh16 2016 , ApJ, 821, 56

    Fausnaugh, M., Denney, K., Barth, A., et al. Fausnaugh16 2016 , ApJ, 821, 56

    2016

  39. [47]

    et al., Fausnaugh17 2017 , Frontiers in Astronomy and Space Sciences, 4, 55

    Fausnaugh, M., Peterson, B., Starkey, D. et al., Fausnaugh17 2017 , Frontiers in Astronomy and Space Sciences, 4, 55

    2017

  40. [48]

    Fernández, J., Chou, D., Pan, Y., Wang, L., Fernandez12 2012 , PASP, 124, 915, 507-512

    2012

  41. [49]

    Gardner, E., and Done, C., GD17 2017 , MNRAS, 470, 3591–3605

    2017

  42. [50]

    Hagen, S., Done, C., and Edelson, R., Hagen24 2024 , MNRAS, 530, 4850–4867

    2024

  43. [51]

    V., Edelson, R., Horne, K., et al

    Hernández Santisteban, J. V., Edelson, R., Horne, K., et al. Hernandez20 2020 , MNRAS, 498, 5399

    2020

  44. [52]

    Honeycutt, R., Honeycutt92 1992 , PASP, 104, 435

    1992

  45. [53]

    Kammoun, E., Papadakis, I., Dovčiak, M., Kammoun21 2021 , MNRAS, 503, 4163–4171

    2021

  46. [54]

    Kasliwal, V., Vogeley, M., and Richards, G., Kasliwal15 2015 , MNRAS, 451, 4328–4345

    2015

  47. [55]

    A., et al

    Kara, E., Mehdipour, M., Kriss, G. A., et al. Kara21 2021 , ApJ, 922, 151

    2021

  48. [56]

    Kelly, B., Bechtold, J., and Siemiginowska, A., Kelly09 2009 , ApJ, 698, 895–910

    2009

  49. [57]

    Korista K., Goad M., KG19 2019 , MNRAS, 489, 5284

    2019

  50. [58]

    et al., Krolik91 1991 , ApJ, 371, 541-562

    Krolik, J., Horne, K., Kallman, T. et al., Krolik91 1991 , ApJ, 371, 541-562

    1991

  51. [59]

    Li14 2014 , ApJ, 786, L6

    Li, Y., Wang, J., Hu C., et al. Li14 2014 , ApJ, 786, L6

    2014

  52. [60]

    Netzer H., Netzer20 2020 , MNRAS, 494, 1611-1621

    2020

  53. [61]

    M., Peterson93 1993 , PASP, 105, 247-268

    Peterson, B. M., Peterson93 1993 , PASP, 105, 247-268

    1993

  54. [62]

    Sergeev05 2005 , ApJ, 622, 129–135

    Sergeev, S., Doroshenko, V., Golubinskiy, Y., et al. Sergeev05 2005 , ApJ, 622, 129–135

    2005

  55. [63]

    Shakura N., Sunyaev R., SS73 1973 , A&A, 500, 33

    1973

  56. [64]

    et al., Shen24 2024 , ApJS, 272, 26

    Shen, Y., Grier, C., Horne, K. et al., Shen24 2024 , ApJS, 272, 26

    2024

  57. [65]

    Simm, T., Salvato, M., Saglia, R., et al., Simm16 2016 , A&A, 585, A129

    2016

  58. [66]

    Starkey, D., Huang, J., Horne, K., and Lin, D., Starkey22 2022 , 519, 2754–2768

    2022

  59. [67]

    Sun20 2020 , ApJ, 891, 178

    Sun, M., Xue, Y., Brandt, W., et al. Sun20 2020 , ApJ, 891, 178

    2020

  60. [68]

    Tamuz, O., Mazeh, T., and Zucker, S., Tamuz05 2005 , MNRAS, 356, 1466-1470

    2005

  61. [69]

    Urry, C., and Padovani, P., UP95 1995 , PASP, 107, 803-845

    1995

Pith tools

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