Pith. sign in

REVIEW 3 major objections 5 minor 58 references

Extreme value distribution for gamma-ray-burst prompt data -- How unexpected was the BOAT event?

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read Using generalized extreme value statistics on 30-day block maxima from the BATSE and Fermi-GBM catalogues, the paper shows that GRB 221009A is grossly inconsistent with the single-population GEV for fluence and peak flux, with median…

desk verdict A genuinely new GEV-based view of the BOAT outlier, with a robust qualitative conclusion but headline return periods that are in-sample and fragile; deserves refereeing, not desk rejection. read the letter →

arxiv 2507.14041 v2 pith:HOBZQ5FF submitted 2025-07-18 astro-ph.HE

classification astro-ph.HE MSC 62G3262F1562P35 PACS 98.70.Rz
keywords gamma-rayburstsGRB221009AextremevaluetheorygeneralizeddistributionFermi-GBMCGRO-BATSEreturnlevelfluence
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 test, without invoking specific emission models, whether the brightest gamma-ray bursts observed by CGRO-BATSE and Fermi-GBM form a single homogeneous population. Using the generalized extreme value (GEV) distribution fitted to 30-day block maxima of fluence, peak flux, and duration, it finds that all three quantities are well described by GEV fits, except that GRB 221009A (the "BOAT") is a dramatic outlier in fluence and peak flux, with median return levels of about one event per millennium and one event per roughly 140 years, respectively. The paper argues that, if the assumptions hold, the BOAT's brightness in fluence and flux is statistically inconsistent with the population that produced all other bright bursts, while its duration is not unusual.

What carries the argument

The machinery is the generalized extreme value (GEV) distribution, a three-parameter family (location $\mu$, scale $\phi$, shape $\xi$) that arises as the limit law for block maxima under very general conditions, analogous to the central limit theorem for sums. The paper takes 30-day blocks from the BATSE and GBM catalogues, records the maximum fluence, peak flux, and $T_{90}$ duration in each block, fits the GEV in a Bayesian framework with improper uniform priors, and uses the fitted law to compute return levels, the value expected to be exceeded once per given period. The shape parameter governs the tail: a negative $\xi$ implies a finite upper limit, and the paper finds that without the BOAT the fluence GEV has a negative shape with an upper boundary near the BOAT's fluence.

What would settle it

Observing any additional GRB within the next few decades with a saturation-corrected fluence comparable to or larger than that of GRB 221009A (around 0.1 erg cm$^{-2}$) would directly contradict the paper's median return level of about one per millennium. A re-analysis using longer blocks (e.g., 90 days or one year) that places the BOAT inside the model's 95% credible region would also undermine the claimed inconsistency.

Watch

Extended reading notes

Core claim

The central claim is that the extreme-value statistics of prompt GRB data are consistent with a single parent population, apart from GRB 221009A. Fitting a GEV distribution to the maxima extracted in equal time slots, the paper finds that the QQ plots for BATSE and Fermi-GBM fluence, peak flux, and duration all follow the model within the 95% credible region, with the notable exception of the BOAT for fluence and peak flux. The median return level for the BOAT's fluence is about 1000 years, and for its peak flux about 140 years; the observed peak flux enters the 95% allowed range only at return periods of about 30 years. Removing the BOAT from the fit yields a fluence GEV with an upper boundary around 0.08-0.09 erg cm$^{-2}$, close to the observed BOAT fluence, while its duration is perfectly consistent with the population. The paper concludes that the BOAT is "grossly inconsistent" with the GEV distribution for fluence and peak flux, and that if one insists it is a typical cosmological GRB, then one of the assumptions, stationarity or a single homogeneous population, must be violated.

Load-bearing premise

The load-bearing premise is that the bright events are drawn from a single, stationary population and that detection efficiency for bright bursts is constant over each mission; if the BOAT belongs to a different population or the rate or efficiency varies, the fitted GEV and the derived return periods no longer apply.

Editorial extensions

If this is right

  • If the BOAT is not a typical GRB, its occurrence does not require the entire bright-end population to be revised; the single-population GEV remains valid for the rest.
  • The fitted fluence upper boundary of about 0.08-0.09 erg cm$^{-2}$, if real, predicts that no physically allowed cosmological GRB should substantially exceed the BOAT's fluence.
  • Return levels give quantitative forecasts: a GRB with peak flux like the BOAT's should be seen roughly once every 140 years, so current monitors are unlikely to catch another in the next few decades.
  • The simultaneous consistency of the duration and inconsistency of fluence and flux supports the view that the BOAT's exceptionality comes from a combination of sustained high flux and spectral hardness, not from an extreme $T_{90}$.

Reading between the lines

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

  • The same block-maxima GEV machinery could be applied to other transient populations, such as fast radio bursts or supernovae, where the question is whether a record event belongs to the same physical class.
  • If the single-population assumption is relaxed, the paper's numbers quantify how rare the BOAT would be in a mixture model: the rate of such events would have to be about one per millennium in fluence and one per roughly 140 years in flux to match.
  • A practical extension would be to incorporate detection-efficiency corrections and mission lifetimes explicitly as covariates in the GEV fit, turning the stationarity assumption into a testable regression.
  • The discrepancy also suggests that searches for gravitational-lensing or narrow-jet explanations are statistically motivated: something must break the single-population assumption for the fluence and flux, and the required rate is low enough that such explanations need not affect the bulk of the population.
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. This paper applies the generalised extreme value (GEV) formalism to 30-day block maxima of fluence, peak flux, and T90 duration drawn from the BATSE and Fermi-GBM GRB catalogues. The central empirical claim is that the GEV distributions describe the bulk of the bright-event population well, but GRB 221009A is a gross outlier in fluence and peak flux while being unremarkable in duration. The paper quotes median return periods of about 1000 years for the fluence and 140 years for the peak flux, and it reports that removing the BOAT from the fit makes the probability of a comparable event 'negligible' on thousands-of-year timescales. The analysis is Bayesian, uses improper priors, and the code and data are made publicly available.

Significance. If the qualitative conclusion is robust, the paper provides a valuable, near-model-independent confirmation that GRB 221009A is an extreme statistical outlier under single-population and stationarity assumptions, complementing earlier estimates based on logN-logS and population modelling. The paper deserves credit for explicitly reporting the leave-one-out sensitivity analysis and for releasing the data and code, which makes the analysis reproducible. The main caveat is quantitative: the specific return-period numbers quoted as headline results are in-sample estimates that are unstable under a simple leave-one-out perturbation, so the central qualitative conclusion is stronger than the specific numerical recurrence rates.

major comments (3)
  1. [Abstract and §5] The headline return periods (~1000 yr for fluence, ~140 yr for peak flux) are in-sample estimates because the GEV parameters are fitted to a sample that includes GRB 221009A, and the GEV fit is strongly influenced by the largest observation. The paper's own leave-one-out result in §4 and Table B.1 shows that excluding the BOAT changes the fluence shape parameter from −0.04±0.03 to −0.12±0.03 and yields a 'negligible' probability of a comparable fluence on timescales of thousands of years. Thus the abstract's 'one event per millennium' claim is not robust to removing the very event it is meant to evaluate. The authors should either present the leave-one-out result as the out-of-sample estimate (which would make the fluence outlier even more extreme) or explicitly state in the abstract and conclusions that the quoted return periods are in-sample and become much longer when the target event is excluded.
  2. [§4] The return-level extrapolation is extremely far beyond the observed data span. With 30-day blocks, the 207-block sample covers about 17 years, so a 1000-year return level corresponds to roughly 12,000 blocks—about a factor of 60 extrapolation. The paper does not demonstrate that the conclusions are stable under different block lengths (e.g., 15-day or 60-day blocks), nor does it provide profile-likelihood or other uncertainty estimates for the return levels that specifically capture the extrapolation uncertainty. Given that the full-sample fluence shape parameter is consistent with zero (−0.04±0.03) while the leave-one-out fit is negative (−0.12±0.03), the tail behaviour is not well pinned down; a block-length sensitivity analysis would strengthen the quantitative claims considerably.
  3. [§4 and §5] The text uses 'the BOAT fluence enters the 95% allowed range at ~150 years' and 'enters the 95% range in ~30 years' without specifying which quantile of the posterior return-level distribution is being compared to the observed value. Because the return-level curves have wide credible intervals, the numerical claims are ambiguous and not directly reproducible from the figures. The authors should state explicitly whether they compare the observed value to the median, the 2.5% lower bound, or another quantile of the return-level distribution.
minor comments (5)
  1. [§4 and §5] GRB 221009A is mislabelled as 'GRB 220109A' in several places (e.g., §4 text and the notes to Table B.1); the same typo appears in the last paragraph of §1 as 'GRB 221009'. These should be corrected to avoid confusion with the actual GRB 220109A.
  2. [Title page] The received and accepted dates ('Received September 15, 1996; accepted March 16, 1997') are clearly erroneous and should be updated or removed.
  3. [§3] The sentence 'Choosing different baselines for the peak flux computation does not change any of the conclusions reported in this paper' is a claim rather than a demonstration; a brief mention of the alternative baseline tried would help the reader judge it.
  4. [Fig. 2] The return-level y-axis in Fig. 2 is labelled in log10 units but the tick labels are not consistently prefixed with log10; please make the axis labelling uniform across panels.
  5. [§5] The statement that a ~1000 s duration corresponds to a return period of 20-30 years in both catalogues would benefit from being tied directly to Fig. 2 or to a table entry, so the reader can verify it.

Circularity Check

1 steps flagged · score 6.0 of 10

Headline return periods for the BOAT are read from a GEV fit that includes the BOAT itself; the paper's own leave-one-out fit makes the probabilities negligible, so the quantitative recurrence claims are in-sample rather than independent predictions.

  1. fitted input called prediction [Abstract; Sect. 4 (return-level paragraph); Sect. 5; Table B.1]
    "Just extrapolating the fit distribution, the median value of the model return level, at the fluence of GRB 220109A, is more than∼ 1000 years ... Removing GRB 220109A the GEV distributions for the fluence and peak flux gives a negligible probability of observing an event of comparable fluence or peak flux for timescales of thousands years. ... The median of the return level for the fluence yields a probability of one event of this fluence occurring in about 1000 years, while for the peak flux the median return value is about 140 years."

    The GEV parameters are estimated from the 207 Fermi-GBM block maxima, a sample that includes GRB 221009A (printed as 'GRB 220109A' in the text) itself. The quoted return periods are read off that same fitted GEV at the BOAT's fluence and peak flux, so the 'prediction' is an in-sample evaluation: the event used to estimate the model is the event whose recurrence probability is reported. The paper's own leave-one-out fit shows the fluence shape parameter changes from -0.04±0.03 to -0.12±0.03 and that without the BOAT a comparable event has 'negligible probability' on thousand-year timescales. The headline ~1000-year and ~140-year values therefore reduce to the parameters that were fitted with the target included; they are not independent out-of-sample predictions.

full rationale

No self-citation or imported-uniqueness pattern is present; the paper is a single-author statistical analysis using public catalogues and standard GEV theory. The only circular element is the headline quantitative claim: the return periods quoted in the Abstract and Sect. 5 are computed from a GEV fit whose 30-day block-maxima sample includes GRB 221009A, and the same fitted distribution is then evaluated at the BOAT's own fluence and peak flux to produce the 'one per millennium' and 'one per century' numbers. This is an in-sample statistic, not an out-of-sample prediction. The paper is transparent about the issue and performs a leave-one-out fit that changes the fluence shape parameter and makes a comparable event's probability 'negligible' on thousand-year timescales, which confirms the qualitative outlier conclusion but also shows the specific quantitative probabilities are not robust to the target's inclusion. Because the central claim that the BOAT is grossly inconsistent with the GEV distribution has independent support from the QQ plot, the raw data display, and the leave-one-out analysis, the circularity is partial rather than total. Score 6 reflects that one or more of the paper's numerical 'predictions' reduce to the fitted model that already contains the predicted event.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central GEV analysis rests on standard extreme-value theory plus three domain assumptions: stationarity, a single homogeneous population, and reliability of the saturation-corrected BOAT measurements. The fitted GEV parameters are free parameters; the 30-day block length is a hand-chosen hyperparameter. No new physical entities are introduced.

free parameters (6)
  • GEV location mu (Fermi-GBM fluence, full) = -4.41 +/- 0.03
    Fitted to 30-day block maxima of GBM fluences (Table B.1); determines return levels in log fluence.
  • GEV log-scale log phi (Fermi-GBM fluence, full) = -0.40 +/- 0.02
    Fitted scale of the extreme-value distribution for GBM fluence (Table B.1).
  • GEV shape xi (Fermi-GBM fluence, full) = -0.04 +/- 0.03
    Fitted tail shape; value near zero makes the tail nearly Gumbel, which drives the about 1000-year return-level extrapolation.
  • GEV shape xi (Fermi-GBM fluence, excluding BOAT) = -0.12 +/- 0.03
    When GRB 221009A is removed, the shape becomes clearly negative, implying a finite upper bound near 0.08-0.09 erg/cm2, and the return probability on thousand-year timescales becomes negligible.
  • GEV shape xi (Fermi-GBM peak flux, full) = 0.00 +0.05/-0.04
    Fitted shape for peak-flux extremes; drives the about 140-year peak-flux return estimate.
  • Block length for extreme-value selection = 30 days
    Chosen by trial-and-improvement as a bias-variance compromise (Sect. 4); not fixed by first principles, and different block lengths would shift the fitted GEV parameters.
assumptions (5)
  • standard math GEV limit theorem applies to block maxima of iid random variables from a continuous distribution (Fisher-Tippett-Gnedenko)
    Invoked in Sect. 2 as the basis for the entire analysis; asymptotic result with regularity conditions.
  • domain assumption Detection probability of bright GRBs is essentially constant over each mission lifetime, and the GRB rate is stationary in time
    Stated in Sect. 1 as one of the two core assumptions enabling the GEV approach; if false, the block-maxima distribution is not stationary and return levels are invalid.
  • domain assumption The observed bright events (extreme values) are drawn from a single homogeneous population of cosmological GRBs
    Sect. 5: consistency with a single population is the hypothesis being tested; the paper explicitly notes that if the BOAT belongs to another population (lensed, narrow-beam, non-cosmological), the derived probabilities do not apply.
  • domain assumption Saturation-corrected fluence and peak flux of GRB 221009A from Lesage et al. (2023) and Frederiks et al. (2023) are reliable
    Adopted in Sect. 3; the peak-flux correction changes the value by an order of magnitude, and errors are not propagated into GEV credible intervals.
  • domain assumption Block maxima from 30-day windows are well approximated by the GEV limit distribution for the sample sizes used
    Sect. 2 acknowledges the choice is not driven by first principles; QQ plots are used as diagnostics, but the approximation quality is assumed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Extreme value distribution for gamma-ray-burst prompt data -- How unexpected was the BOAT event?." pith.science (2026). https://pith.science/paper/HOBZQ5FF

@misc{pith2026250714041,
  author       = {Pith},
  title        = {Pith review of: Extreme value distribution for gamma-ray-burst prompt data -- How unexpected was the BOAT event?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HOBZQ5FF}},
  note         = {Machine review of arXiv:2507.14041}
}
read the original abstract

Gamma-Ray Bursts (GRBs) are known to be unpredictable in time and position. A few (observationally) exceptional events have been observed, as GRB221009A that stands out for its fluence and peak flux, being orders of magnitude higher than what measured so far. Analyzing the observed fluence, peak flux or duration distributions typically requires one to assume some scenarios, and the consistency of the observed data with the predictions turns out to be an important model diagnostic. However, it is also of interest to model these distributions using general statistical properties that do not rely on specific model assumptions, allowing one to derive inferences only based on the consistency of the observed distributions with the hypothesis of one single population of events that generate them. We obtained fluences, peak fluxes and durations from the catalogues of GRBs observed by the CGRO-BATSE and Fermi-GBM instruments. We selected the extreme values in slots of equal duration and modelled their distributions by the generalized extreme value (GEV) formalism. The GEV distribution is a limit distribution naturally arising when the number of observations is large and is essentially independent of the phenomena producing the observed data. The distributions of extreme values for fluences, peak fluxes and durations are consistent with being extracted from a single population of events but the fluence and peak flux recorded for GRB221009A constitutes a striking exception. The probability to observe such an event, assuming it is a cosmological GRB, is low, with a median value of about one event per millennium for the fluence and about one event per century for the peak flux.

Figures

Figures reproduced from arXiv: 2507.14041 by the authors.

Figure 1
Figure 1. QQ plots for GEV fit of extreme values extracted from the duration (bottom), peak flux (middle), and fluence (top) of the CGRO-BATSE catalogue (left) and Fermi-GBM catalogue (right). The 95% credible region is also shown. GRB 221009A (red symbol) is clearly an outlier for the fluence and peak flux, but it is not for the duration. The insets show the results of the fit excluding GRB 221009A from the data set. GRB 230… view at source ↗
Figure 2
Figure 2. Return plots for GEV fit of extreme values extracted from the duration (bottom), peak flux (middle) and fluence (top) of the GRO￾BATSE catalogue (left), and Fermi-GBM catalogue (right). The 95% credible region is also shown. The model cannot interpret the return level of GRB 221009A for the fluence and peak flux while this event is part of the general distribution for duration. The insets show the results of the fit… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 51 canonical work pages

  1. [1]

    P., Anand , S., et al

    Ahumada , T., Singer , L. P., Anand , S., et al. 2021, Nature Astronomy, 5, 917

  2. [2]

    2006, , 372, 233

    Amati , L. 2006, , 372, 233

  3. [3]

    2025, , 980, 241

    Atteia , J.-L., Bouchet , L., Dezalay , J.-P., et al. 2025, , 980, 241

  4. [4]

    2004, Statistics of Extremes: Theory and Applications (Wiley), pagination: 522

    Beirlant, J., Goegebeur, Y., Segers, J., & Teugels, J. 2004, Statistics of Extremes: Theory and Applications (Wiley), pagination: 522

  5. [5]

    Bezanson, J., Edelman, A., Karpinski, S., & Shah, V. B. 2017, SIAM R eview, 59, 65

  6. [6]

    N., Meegan , C

    Bhat , P. N., Meegan , C. A., von Kienlin , A., et al. 2016, VizieR Online Data Catalog: The third Fermi/GBM GRB catalog (6yr) (Bhat+, 2016) , VizieR On-line Data Catalog: J/ApJS/223/28. Originally published in: 2016ApJS..223...28B

  7. [7]

    Bloom , J. S. 2022, Research Notes of the American Astronomical Society, 6, 220

  8. [8]

    A new long gamma-ray burst formation pathway at solar metallicity

    Briel , M. M., Fragos , T., Salafia , O. S., et al. 2025, arXiv e-prints, arXiv:2502.09187

Show all 58 references
  1. [9]

    S., Paciesas , W

    Briggs , M. S., Paciesas , W. S., Pendleton , G. N., et al. 1996, , 459, 40

  2. [10]

    2011, Handbook of Markov chain Monte Carlo (CRC press)

    Brooks, S., Gelman, A., Jones, G., & Meng, X.-L. 2011, Handbook of Markov chain Monte Carlo (CRC press)

  3. [11]

    2023, , 946, L31

    Burns , E., Svinkin , D., Fenimore , E., et al. 2023, , 946, L31

  4. [12]

    2001, An introduction to statistical modeling of extreme values, Springer Series in Statistics (London: Springer-Verlag)

    Coles, S. 2001, An introduction to statistical modeling of extreme values, Springer Series in Statistics (London: Springer-Verlag)

  5. [13]

    G., Lenart , A

    Dainotti , M. G., Lenart , A. ., Chraya , A., et al. 2023, , 518, 2201

  6. [14]

    & Krumbiegel, J

    Danisch, S. & Krumbiegel, J. 2021, Journal of Open Source Software, 6, 3349

  7. [15]

    2020, Journal of Open Source Software, 5, 2673

    Datseris, G., Isensee, J., Pech, S., & Gál, T. 2020, Journal of Open Source Software, 5, 2673

  8. [16]

    2023, , 954, L29

    Dichiara , S., Tsang , D., Troja , E., et al. 2023, , 954, L29

  9. [17]

    E., Epstein , R

    Fenimore , E. E., Epstein , R. I., Ho , C., et al. 1993, , 366, 40

  10. [18]

    Finke , J. D. & Razzaque , S. 2024, , 975, 70

  11. [19]

    2011, A History of the Central Limit Theorem : From Classical to Modern Probability Theory (New York, NY: Springer New York)

    Fischer, H. 2011, A History of the Central Limit Theorem : From Classical to Modern Probability Theory (New York, NY: Springer New York)

  12. [20]

    L., et al

    Frederiks , D., Svinkin , D., Lysenko , A. L., et al. 2023, , 949, L7

  13. [21]

    2024, arXiv e-prints, arXiv:2412.21175

    Galanti , G., Nava , L., Roncadelli , M., Tavecchio , F., & Bonnoli , G. 2024, arXiv e-prints, arXiv:2412.21175

  14. [22]

    2025, arXiv e-prints, arXiv:2502.03453

    Galanti , G., Roncadelli , M., Bonnoli , G., Nava , L., & Tavecchio , F. 2025, arXiv e-prints, arXiv:2502.03453

  15. [23]

    2012, , 420, 483

    Ghirlanda , G., Nava , L., Ghisellini , G., et al. 2012, , 420, 483

  16. [24]

    D., Mallozzi , R

    Goldstein , A., Preece , R. D., Mallozzi , R. S., et al. 2013, , 208, 21

  17. [25]

    D., Caceres , G

    Gondhalekar , Y., Feigelson , E. D., Caceres , G. A., Montalto , M., & Saha , S. 2023, , 959, L16

  18. [26]

    2014, , 211, 12

    Gruber , D., Goldstein , A., Weller von Ahlefeld , V., et al. 2014, , 211, 12

  19. [27]

    2024, , 534, 173

    Heather , C., Chantavat , T., Chongchitnan , S., & Silk , J. 2024, , 534, 173

  20. [28]

    Hoffman, M. D. & Gelman, A. 2014, Journal of Machine Learning Research, 15, 1593

  21. [29]

    2024, Journal of Statistical Software, 109, 1–35

    Jalbert, J., Farmer, M., Gobeil, G., & Roy, P. 2024, Journal of Statistical Software, 109, 1–35

  22. [30]

    D., Briggs , M

    Kaneko , Y., Preece , R. D., Briggs , M. S., et al. 2006, , 166, 298

  23. [31]

    W., Strong , I

    Klebesadel , R. W., Strong , I. B., & Olson , R. A. 1973, , 182, L85

  24. [32]

    A., Fishman , G

    Kouveliotou , C., Meegan , C. A., Fishman , G. J., et al. 1993, , 413, L101

  25. [33]

    2023, , 949, L4

    Lan , L., Gao , H., Li , A., et al. 2023, , 949, L4

  26. [34]

    S., et al

    Lesage , S., Veres , P., Briggs , M. S., et al. 2023, , 952, L42

  27. [35]

    J., Gompertz , B

    Levan , A. J., Gompertz , B. P., Salafia , O. S., et al. 2024, , 626, 737

  28. [36]

    J., Tanvir , N

    Levan , A. J., Tanvir , N. R., Starling , R. L. C., et al. 2014, , 781, 13

  29. [37]

    B., Levan , A

    Malesani , D. B., Levan , A. J., Izzo , L., et al. 2023, arXiv e-prints, arXiv:2302.07891

  30. [38]

    2024, arXiv e-prints, arXiv:2410.18131

    Navia , C., Oliveira , M., Felicio , B., & Nepomuceno , A. 2024, arXiv e-prints, arXiv:2410.18131

  31. [39]

    Northrop, P. J. & Attalides, N. 2016, Statistica Sinica, 26, 721

  32. [40]

    2023, Science Advances, 9, eadi1405

    O'Connor , B., Troja , E., Ryan , G., et al. 2023, Science Advances, 9, eadi1405

  33. [41]

    2021, , 913, 60

    Poolakkil , S., Preece , R., Fletcher , C., et al. 2021, , 913, 60

  34. [42]

    2013, , 763, 15

    Qin , Y., Liang , E.-W., Liang , Y.-F., et al. 2013, , 763, 15

  35. [43]

    & Szapudi , I

    Repp , A. & Szapudi , I. 2018, , 473, 3598

  36. [44]

    D., et al

    Salvaterra , R., Campana , S., Vergani , S. D., et al. 2012, , 749, 68

  37. [45]

    2014, , 440, 2099

    S \"u veges , M. 2014, , 440, 2099

  38. [46]

    2023, PairPlots.jl Beautiful and flexible visualizations of high dimensional data, https://sefffal.github.io/PairPlots.jl/dev

    Thompson, W. 2023, PairPlots.jl Beautiful and flexible visualizations of high dimensional data, https://sefffal.github.io/PairPlots.jl/dev

  39. [47]

    2017, , 850, 161

    Tsvetkova , A., Frederiks , D., Golenetskii , S., et al. 2017, , 850, 161

  40. [48]

    A., Paciesas , W

    von Kienlin , A., Meegan , C. A., Paciesas , W. S., et al. 2014, , 211, 13

  41. [49]

    A., Paciesas , W

    von Kienlin , A., Meegan , C. A., Paciesas , W. S., et al. 2020, , 893, 46

  42. [50]

    C., Ettori , S., & Moscardini , L

    Waizmann , J. C., Ettori , S., & Moscardini , L. 2012, , 422, 3554

  43. [51]

    A., Kennea , J

    Williams , M. A., Kennea , J. A., Dichiara , S., et al. 2023, , 946, L24

  44. [52]

    2022, , 612, 232

    Yang , J., Ai , S., Zhang , B.-B., et al. 2022, , 612, 232

  45. [53]

    2024, , 626, 742

    Yang , Y.-H., Troja , E., O'Connor , B., et al. 2024, , 626, 742

  46. [54]

    X., Wang , C

    Yi , S. X., Wang , C. W., Shao , X., et al. 2025, , 985, 239

  47. [55]

    2004, , 609, 935

    Yonetoku , D., Murakami , T., Nakamura , T., et al. 2004, , 609, 935

  48. [56]

    2018, The Physics of Gamma-Ray Bursts (Cambridge University Press)

    Zhang, B. 2018, The Physics of Gamma-Ray Bursts (Cambridge University Press)

  49. [57]

    B., Liu , Z

    Zhang , B. B., Liu , Z. K., Peng , Z. K., et al. 2021, Nature Astronomy, 5, 911

  50. [58]

    2024, Journal of Astrophysics and Astronomy, 45, 14

    Zhang , R., Chen , Y.-Q., Zeng , S.-G., et al. 2024, Journal of Astrophysics and Astronomy, 45, 14

Pith tools

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