How Distance Affects GRB Prompt Emission Measurements

Amy Y. Lien; Craig B. Markwardt; Michael J. Moss; S. Bradley Cenko; Sylvain Guiriec

REVIEW 4 major objections 6 minor 3 cited by

High-redshift gamma-ray bursts measured by Swift/BAT have durations underestimated by up to ~10x and fluences by ~2 because faint emission sinks into background noise, and this bias, not new physics, drives the observed redshift trends.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-03 05:05 UTC pith:R6XXQ3H4

load-bearing objection A careful, reproducible simulation study that convincingly shows the tip-of-the-iceberg bias for Swift/BAT, but the stronger claim of full consistency with observed high-z bursts leans on an unexplained factor-2 calibration and a peak-flux mismatch. the 4 major comments →

arxiv 2602.03032 v1 pith:R6XXQ3H4 submitted 2026-02-03 astro-ph.HE

How Distance Affects GRB Prompt Emission Measurements

Michael J. Moss , Amy Y. Lien , S. Bradley Cenko , Sylvain Guiriec , Craig B. Markwardt This is my paper

classification astro-ph.HE PACS 98.70.Rz

keywords gamma-ray burstsprompt emissionredshiftT90 durationfluencetip-of-the-iceberg effectSwift/BATselection bias

verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper argues that the apparent shortening of gamma-ray burst durations with increasing redshift is an observational artifact, not a real change in the bursts. The authors take 26 bright bursts at z<1 and simulate how the Swift/BAT instrument would observe them at redshifts up to 15, folding in time dilation, spectral shifting, instrument response, and realistic background noise. They find that faint emission progressively sinks into the background—the 'tip-of-the-iceberg' effect—so measured durations (T90) come out up to an order of magnitude too short and fluences about a factor of two too low. The simulated distant bursts match 72 real bursts observed at z>3 in duration and fluence, implying the real ones carry the same underestimates. If the paper is right, no new physics or population evolution is needed to explain the observed duration-redshift trend, and the inferred energies of distant bursts are systematically low.

Core claim

The paper's central claim is that the 'tip-of-the-iceberg' effect—faint emission slipping below detector background as signals weaken—quantitatively accounts for how GRB prompt-emission measurements change with redshift. The authors take 26 bright bursts at z<1, rescale each to redshifts up to 15 with proper time dilation, k-correction, luminosity distance, and the Swift/BAT instrument response, add realistic background fluctuations, and re-measure durations with the Bayesian block algorithm. Measured durations (T90, the interval containing 5% to 95% of the burst counts) come out shorter than the true, time-dilated value—sometimes by an order of magnitude—and fluences run low by up to a fact

What carries the argument

The carrying mechanism is the tip-of-the-iceberg effect: as signal-to-noise falls, dim pulses and faint tails of a burst's light curve drop below the detector background and become unrecoverable, so measured durations and fluences are truncated. The quantitative machinery is a simulation pipeline that rescales each burst's observed light curve and best-fit 1-second peak cut-off power-law spectrum to higher redshift using luminosity distance and k-correction, folds the redshifted spectrum through the Swift/BAT response (scaled by a factor of 2 to keep the simulation consistent with the input observations), adds background fluctuations drawn from an empirical distribution of BAT background var

Load-bearing premise

The load-bearing premise is the simulation's calibration: each burst's best-fit 1-second peak spectrum is applied to the entire burst, and the folded spectra are scaled by a factor of 2 to keep the simulation consistent with the input observations; if real spectra are softer at the start and end of emission, or if that scaling is miscalibrated, the simulated signal loss—and with it the reported bias factors—would shift.

What would settle it

Re-run the same pipeline on the same 26 bursts using time-resolved spectra instead of the single 1-second peak spectrum. Because the paper notes burst spectra are typically softer at the start and end of emission, this should lose signal faster; if the resulting simulated high-redshift duration and fluence distributions shift enough that the two-sample KS-tests against the observed high-z sample fall out of agreement (or if the estimated bias factors change by more than the quoted ~10x in duration and ~2 in fluence), the paper's quantitative claim fails.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

Measured T90 durations of high-redshift Swift/BAT GRBs should be read as lower limits: the real, time-dilated durations can be about an order of magnitude longer than reported.
Isotropic gamma-ray energies (Eγ,iso) of distant bursts are systematically underestimated by up to a factor of ~2, so the true Eγ,iso distribution extends to higher energies than previously inferred.
Some intrinsically long GRBs will be misclassified as short GRBs at high redshift purely from this observational bias, slightly contaminating short-burst samples.
The average duration and fluence trends with redshift—flattening and then declining—are explained by the bias alone; no evolution of the GRB population or new emission physics is required.
Measured gamma-ray duration is not a reliable measure of how long a GRB's central engine stays active; recovering true durations needs more sensitive detectors or different estimators.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Because the bias factors depend on the factor-of-2 spectral scaling and the single-peak-spectrum assumption, the specific numbers should be read as calibration-dependent even though the direction of the bias is not; removing or explaining that scaling is the most direct way to firm them up.
The strong agreement with the pre-2012 observed subsample is a hint that the real high-z sample is mission-epoch dependent; comparisons run on bursts collected in different Swift/BAT eras may reach different conclusions about duration evolution, which future samples can test directly.
The same pipeline can be run in reverse to generate bias-vs-sensitivity curves for future instruments: for a chosen detector sensitivity, it predicts how much of a burst's true duration and fluence would be recovered at a given redshift, a practical design target.
If more sensitive X-ray monitors continue to record emission outlasting the gamma-ray signal, the duration gap is an independent, observable measure of the signal the tip-of-the-iceberg effect removes at a given sensitivity—a direct empirical check on these simulations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Desk Editor's Note

A careful, reproducible simulation study that convincingly shows the tip-of-the-iceberg bias for Swift/BAT, but the stronger claim of full consistency with observed high-z bursts leans on an unexplained factor-2 calibration and a peak-flux mismatch.

read the letter

Quick take: this is a solid, clearly written simulation study showing that distance alone can shorten measured T90 and reduce fluence for Swift/BAT bursts at high z. The qualitative result is robust and worth having. The stronger claim—that observed high-z bursts need no additional physics—is plausible but not nailed down, because two things in the method are softer than the abstract implies.

What's new: earlier work (Bonnell, Kocevski & Petrosian, Littlejohns) simulated similar biases, but this is the first to do it with real light curves from 26 bright z<1 bursts, proper k-corrections, full BAT response folding, and Bayesian-block measurement. The code and data are public, and the authors are transparent about their choices. The long-to-short GRB result is genuinely interesting and directly relevant to how we classify bursts.

The main soft spot is the factor-2 scaling of folded spectra. In Section 2 they say they "needed to scale" the folded spectra by 2 to stay self-consistent with input observations, but there is no derivation, no physical justification, no uncertainty. This factor sets the count rate for every simulated burst and therefore drives the SNR ladder that produces the bias. If the true factor is 1 or 4, the simulated high-z distributions shift, and the claimed consistency with observed durations and fluences could change. A referee should ask for the derivation or, at minimum, a sensitivity scan over this factor.

The second issue is the peak-flux mismatch. The simulated high-z subsamples never match the observed peak-flux distribution (0% within 1σ). That means the synthetic bursts are systematically fainter than real high-z bursts, so the agreement in durations and fluences may be partly coincidental—fainter simulated bursts suffer more bias, which could bring them into line with observed (also biased) values. The paper mentions possible causes (the Ep cut, Yonetoku relation) but doesn't resolve the concern.

Third, the simulated high-z sample is drawn from all z>3 simulations and is not matched to the redshift distribution of the observed sample. The KS non-rejection is weak evidence anyway; matching the redshift distributions would make the comparison cleaner. Also a minor typo: the text says 73 observed bursts in one place, but the table has 72.

These problems don't sink the qualitative conclusion—high-z durations and fluences are probably underestimated, and the bias likely contributes to the observed duration-redshift trend. But the "no additional physics" claim is only weakly supported. The paper is honest about several of these caveats, which I appreciate.

Who should read it: anyone using T90, fluence, or Eiso in GRB correlation studies, and anyone interpreting duration-redshift trends. It deserves a real referee. I'd send it to peer review with a request for sensitivity analysis on the factor-2 and a redshift-matched comparison. I'd cite it for the bias quantification, with caution on the strong conclusion.

Referee Report

4 major / 6 minor

Summary. This paper investigates how the observed durations and fluences of Swift/BAT GRB prompt emission change as the source distance increases. The authors select 26 bright z<1 GRBs, forward-model them to redshifts up to z=15 using the public simmes package: each observed light curve is redshifted, the 1-second peak CPL spectrum is k-corrected and folded through the BAT response, background fluctuations are added, and Bayesian blocks are used to measure T90 and fluence. They find that measured durations and fluences are increasingly underestimated at high redshift, occasionally turning long GRBs into short GRBs, and that the simulated z>3 sample is KS-compatible with 72 observed high-z GRBs in T90 and fluence but not in peak flux. The paper concludes that high-z Swift/BAT durations and fluences are likely underestimated by factors of ~several tens and ~2, and that the apparent evolution of these quantities with redshift is consistent with the tip-of-the-iceberg effect.

Significance. If the conclusion holds, the paper would have a significant impact on GRB population studies, especially on claims about the redshift evolution of prompt emission properties and on estimates of E_gamma,iso. The work has clear strengths: it starts from real observed bursts, includes cosmological k-corrections and the BAT response, makes its simulation code and data publicly available, and directly compares synthetic and observed high-z samples. However, the quantitative claims depend on several modeling assumptions that are either unexplained or not fully validated, most importantly the factor-of-2 scaling of folded spectra and the representativeness of the simulated high-z sample. The qualitative tip-of-the-iceberg bias is robust, but the specific underestimate factors and the compatibility claim need additional support before the central quantitative conclusions can be accepted.

major comments (4)

[Section 2 (Methods), folded-spectrum scaling] The statement 'We found that we needed to scale the folded spectra by a factor of 2 to ensure that the simulations would remain self consistent with the input observations' is a load-bearing calibration with no derivation. The folded count flux sets the normalization of every mock light curve, which controls the Bayesian-block SNR and the redshift at which durations and fluences begin to drop. A factor of 2 therefore directly shifts the tip-of-the-iceberg transition and the synthetic z>3 distributions used in the KS tests of Section 4. Please (i) show why this factor is needed, e.g., by comparing predicted and observed count rates for the input bursts; (ii) test the sensitivity of the final results to factor values of 1 and 4; and (iii) propagate the uncertainty in this factor into the quoted underestimate factors. Without this, the abstract's quantitative claims are not established.
[Section 2 (Methods), 1-second peak spectrum assumption] Using the best-fit 1-second peak CPL spectrum for the entire burst is a strong assumption. The authors state that time-resolved spectra did not significantly change their interpretations and that the approximation is conservative, but no supporting analysis is shown. Since the paper quotes specific factors ('several tens' for durations, '~2' for fluences), the effect of spectral evolution on the simulated high-z measurements should be quantified. Please provide the time-resolved comparison or, at minimum, show how realistic variations in photon index and Ep across each burst alter the simulated T90 and fluence distributions.
[Section 4, comparison to observed high-z sample] The synthetic high-z sample is constructed by simulating each low-z burst out to z=15 and then drawing from all simulations with z>3, without weighting by the redshift distribution of the observed high-z sample (Table 5). The observed sample is concentrated at z≈3–5, while the simulation grid extends to much higher z. Because the tip-of-the-iceberg effect strengthens with z, an unweighted synthetic sample may overrepresent the most biased events, making it easier to achieve KS compatibility with the observed durations and fluences. Please repeat the comparison with simulated redshifts drawn from the observed high-z redshift distribution, or show that the results are unchanged under such weighting.
[Section 4, Table 1 and Figure 9] The KS tests show a significant peak-flux mismatch: 0% of simulation subsamples fall within 1σ and only 28.4% within 3σ of the observed high-z peak-flux distribution. Since peak flux sets the SNR, the simulated bursts are systematically fainter than the observed ones and may suffer more severe duration/fluence underestimation. The apparent agreement in durations and fluences could therefore reflect a mismatch in the SNR regime rather than a faithful reproduction of the observed high-z population. The pre-2012 subsample improves agreement, but Appendix A shows that the post-2012 discrepancy is not fully understood. Please address whether a peak-flux-matched subsample changes the duration/fluence comparison, or otherwise explain how the peak-flux mismatch does not compromise the central conclusion.

minor comments (6)

[Abstract and Section 4] The number of observed high-z bursts is given as 72 in the abstract and Table 5, but Section 4 says '73 being the same number of bursts in our observed high-z sample.' Please correct the inconsistency.
[Section 2 (Methods)] The description of normalizing the light curves by the fluence in the 1-second peak interval is unclear. Please specify the exact procedure and how the dimensionless light curve is converted back to count rates.
[Section 2 (Methods)] Please specify how the 50 redshift values are chosen (linear, log-spaced, etc.) and what 'varying the count flux' means in the 1000 realizations per redshift.
[Figure 6d and Section 2] The Bayesian-block threshold in the Figure 6d caption is written as A = sigma sqrt(2 log N) T, while Section 2 defines A = sigma Delta t sqrt(2 log n). Reconcile the notation.
[Appendix A] The p-value to significance conversions (e.g., 'p=0.05 = 2σ') are not standard; please define the conversion used.
[Section 2, background variance PDF] The FRED function fit to the background-variance distribution is described as a probability distribution function, but its normalization and bounds are not stated. Please clarify how the fitted function is normalized and sampled.

Circularity Check

0 steps flagged

No circularity: the simulations are forward models from observed low-z bursts to synthetic high-z bursts, with an external KS-test comparison to real high-z data; the factor-2 scaling is a calibration/uncertainty, not a definitional reduction.

full rationale

The derivation chain is: (1) take 26 observed low-z Swift/BAT GRBs; (2) redshift their observed light curves and best-fit CPL spectra using Eqs. 1-3, including k-corrections and luminosity distance; (3) fold through the BAT response, add empirically sampled background fluctuations, and measure durations/fluences with Bayesian blocks; (4) compare the resulting synthetic z>3 distributions to an independent sample of 72 observed z>3 bursts via two-sample KS tests. The predicted tip-of-the-iceberg underestimation is an emergent property of this forward simulation, not an input. No fitted parameter is renamed as a prediction: the only ad hoc element is the statement 'We found that we needed to scale the folded spectra by a factor of 2 to ensure that the simulations would remain self consistent with the input observations,' which is a calibration of the mock count flux to the input observations and is a modeling uncertainty/correctness risk, not a circular reduction—the high-z observed sample is not used to tune this factor, and the conclusion could in principle fail the KS comparison, as it indeed does for peak flux. The self-citation (Moss et al. 2022) is background context on observational biases and is not load-bearing. No uniqueness theorem, ansatz, or definitional identity is imported to force the result. Therefore no specific circular step can be exhibited.

Axiom & Free-Parameter Ledger

3 free parameters · 4 axioms · 0 invented entities

The central claim rests on two fitted/calibrated quantities (factor-2 spectrum scale, peak-flux sample cut, background-variance FRED fit) and several domain assumptions (constant spectrum, standard cosmology, Bayesian-block reliability, representativeness of low-z templates). No new entities are introduced.

free parameters (3)

Spectrum scale factor = 2
Folded spectra are scaled by factor 2 to make simulations self-consistent with input observations; not independently justified.
Sample peak-flux cut = 2x Bayesian block sensitivity limit at z=3
Hand-chosen threshold to ensure low-z bursts remain detectable at z=3; affects sample composition.
Background variance FRED fit = Fit to Swift/BAT background variance distribution
Used to randomly draw background fluctuations; parameters fit to all BAT GRBs.

axioms (4)

domain assumption GRB spectrum is constant and equal to the best-fit 1-s peak CPL during the whole burst
Invoked in Section 2; if false, signal loss with redshift is miscalibrated.
domain assumption Standard LCDM cosmology with Omega_m=0.3, Omega_L=0.7, H0=67.4
Used for luminosity distances and k-corrections.
domain assumption Bayesian block algorithm correctly identifies source emission intervals in mask-weighted light curves with known background variance
Duration/fluence measurements depend on this; no independent trigger validation.
domain assumption Observed low-z GRBs are representative templates for high-z GRBs after cosmological scaling
Underpins comparison to observed high-z sample; no intrinsic evolution assumed.

pith-pipeline@v1.3.0-alltime-deepseek · 19743 in / 11517 out tokens · 93049 ms · 2026-08-03T05:05:26.990663+00:00 · methodology

0 comments

read the original abstract

We investigated how Gamma-Ray Burst (GRB) prompt emission measurements are affected by increasing distance to the source. We selected a sample of 26 bright GRBs with measured redshifts $z<1$ observed by the Burst Alert Telescope (BAT) on board the Neil Gehrels Swift Observatory (Swift) and simulated what BAT would have observed if the GRBs were at larger redshifts. We measured the durations of the simulated gamma-ray signals using a Bayesian block approach and calculated the enclosed fluences and peak fluxes. As expected, we found that almost all durations (fluences) measured for simulated high-$z$ GRBs were shorter (less) than their true durations (energies) due to low signal-to-noise ratio emission becoming completely dominated by background, i.e., the ``tip-of-the-iceberg'' effect. This effect strongly depends on the profile and intensity of the source light curve. Due to the uniqueness of GRB light curves, there is no common behavior in the evolution of measured durations with redshift. We compared our synthetic high-$z$ (i.e., $z>3$) GRBs to a sample of 72 observed high-$z$ bursts and found that the two samples were not inconsistent with being drawn from the same underlying population. We conclude that: (i) prompt emission durations (fluences) of high-$z$ GRBs observed by Swift/BAT are most likely underestimations, sometimes by factors of $\sim$several tens ($\sim2$), and (ii) changes in the average GRB prompt emission duration and fluence with increasing redshift are consistent with the tip-of-the-iceberg effect.

Figures

Figures reproduced from arXiv: 2602.03032 by Amy Y. Lien, Craig B. Markwardt, Michael J. Moss, S. Bradley Cenko, Sylvain Guiriec.

**Figure 1.** Figure 1: T90 measurements of Swift/BAT GRBs as a function of their measured redshifts. The blue and orange points are GRBs with a measured prompt duration of T90 > 2 s and T90 < 2 s, i.e., long and short GRBs, respectively. The lower and upper violet dashed lines are ∝ (1 + z)T for rest frame durations of T = 25 s and 100 s, respectively. The black line is the weighted mean of the log durations of LGRBs in each se… view at source ↗

**Figure 2.** Figure 2: Work flow schematic of the simulations and measurements performed in this work. ple would be measurable out to at least z = 3 (see yellow region in [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Figure 3: Peak flux measurements of Swift/BAT GRBs as a function of their measured redshifts. Blue points are all Swift/BAT GRBs. The red stars and green triangles indicate GRBs in the low-z and high-z samples of this work, respectively (see Tables 4 and 5). The horizontal gray line indicates the theoretical Bayesian block sensitivity limit, the dashed line is twice that limit (see text for details). For reference,… view at source ↗

**Figure 4.** Figure 4: We compiled the 1-second time-bin background variances from the mask-weighted light curves of all Swift/BAT GRBs (blue) and fit a FRED function to the distribution to create a probability distribution function (orange). The background variances were calculated within the 50 second intervals preceding and following the emission of each burst. We sampled this distribution during our simulations to determi… view at source ↗

**Figure 5.** Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗

**Figure 6.** Figure 6: a: The 15 - 150 keV Light curve of GRB 050525A simulated at increasing redshifts (indicated by lighter shades of blue). b: The fraction of simulations able to be measured by the Bayesian block algorithm in each redshift bin (1, 000 simulations were performed in each z bin). c: Density plot of the measured T90 at increasing redshifts for mock GRBs generated from GRB 050525A data. The black dotted line indic… view at source ↗

**Figure 7.** Figure 7: a: The 15 - 150 keV light curve for GRB 111228A. The orange shaded region indicates the T90 measured by the automated BAT analysis pipeline. Panels b, c, and d are described in [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗

**Figure 8.** Figure 8: Top: The 15 - 150 keV light curve for GRB 120311A as observed by Swift/BAT. The orange shaded region indicates the T90 measured by the automated BAT analysis pipeline. Middle: Same description as Figure 6b, but for simulations using GRB 120311A as the input GRB. The horizontal, black line indicates the traditional separation between long and short GRBs (i.e., 2 seconds). This is an example of a long GRB … view at source ↗

**Figure 9.** Figure 9: Top Row: Cumulative distributions of the T90 (left), fluence (center ), and peak flux (right) measurements for the observed high-z sample (solid, blue) and the simulated high-z sample (dashed orange). The orange shaded regions indicate the range of cumulative distributions generated from 1,000 trials of randomly sampling 73 simulation results from our entire simulated high-z sample (i.e., the same number a… view at source ↗

**Figure 10.** Figure 10: Same description as for [PITH_FULL_IMAGE:figures/full_fig_p012_10.png] view at source ↗

**Figure 11.** Figure 11: Displayed are the cumulative distributions of the measured durations (Top), fluences (Center ), and peak fluxes (Bottom) for Swift/BAT GRBs observed before 2012-01-01 (cyan) and those observed after (olive). The p-values obtained from KS-tests for these distributions can be found in Table We separated our observed high-z GRB sample into GRBs observed before and after 2012-01-01, however, that date was sim… view at source ↗

**Figure 12.** Figure 12: Displayed are the 1-second peak fluxes for a: all Swift/BAT GRBs, b: only GRBs with redshifts, c: only GRBs with redshifts z > 3, and d: only GRBs without redshifts as a function of their observation date. The average peak flux in 6-month bins is shown with the solid lines in each plot. The thin purple line indicates the 2005 average peak flux [PITH_FULL_IMAGE:figures/full_fig_p016_12.png] view at source ↗

**Figure 13.** Figure 13: Average number of enabled/active detectors on the Swift/BAT detector plane. The total number of detectors at launch was 32,768. In early 2020, BAT recovered a significant number of detectors upon a spontaneous reboot. where N is the number of detectors active on the detector plane (G. K. Skinner 2008; W. H. Baumgartner et al. 2013). The ratio between the number of active detectors on BAT at the beginning … view at source ↗

**Figure 14.** Figure 14: Left: The number of redshift measurements obtained for Swift/BAT GRBs as a function of date since mission launch. All Swift/BAT GRBs are shown in gray, Swift/BAT GRBs with redshift in blue, and GRBs with redshifts z > 3 in orange. Middle: The average redshift measured for Swift/BAT GRBs as a function of their observation date. Swift/BAT GRBs are displayed in blue point. The black line indicates the averag… view at source ↗

discussion (0)

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

EP260321a/SN 2026gzf: The Faintest Shock Breakout Associated with a Broad-Lined Supernova
astro-ph.HE 2026-06 unverdicted novelty 7.0

EP260321a is the faintest observed shock breakout tied to a broad-lined Type Ic supernova, interpreted as a choked weak outflow from a stripped star.
4th TDAMM Workshop White Paper
astro-ph.HE 2026-07 conditional novelty 4.0

A community white paper arguing TDAMM science is becoming follow-up-limited, and proposing pre-negotiated community observing plans with infrastructure, policy, and capability recommendations.
EP260321a/SN 2026gzf: The Faintest Shock Breakout Associated with a Broad-Lined Supernova
astro-ph.HE 2026-06 unverdicted novelty 4.0

EP260321a is identified as the faintest shock breakout X-ray transient associated with broad-lined Ic supernova SN 2026gzf, interpreted as originating from a mildly relativistic weak outflow choked inside the progenitor star.

Reference graph

Works this paper leans on

6 extracted references · 1 linked inside Pith · cited by 2 Pith papers

[1]

P., Abbott, R., Abbott, T

Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2017, ApJL, 848, L13, doi: 10.3847/2041-8213/aa920c Ahumada, T., Singer, L. P., Anand, S., et al. 2021, Nature Astronomy, 5, 917, doi: 10.1038/s41550-021-01428-7 Ajello, M., Arimoto, M., Axelsson, M., et al. 2019, ApJ, 878, 52, doi: 10.3847/1538-4357/ab1d4e Astropy Collaboration, Robitaille, T. P., Tollerud...

Pith/arXiv arXiv doi:10.48550/arxiv.2403.13126 2017
[2]

Looking only at Swift/BAT GRBs that have measured redshifts, it seems this peak-flux increase may be more pronounced (see Fig. 12b). For GRBs with redshiftsz >3, this bias does not seem to be present, but the sample size is quite low, especially so after 2016 (see Fig. 12c). But even for GRBs without any redshifts, there seems to be a slight increase in t...

2016
[3]

Naively, since these are all high-zGRBs, these measurements are drawn from the same underlying distribution and a KS-Test should findp≥0.32 to indicate this fact

We can see that duration measurements of the pre-2012 and post-2012 samples are consistent with one another to within 1σ(i.e.,p= 0.61), however, the fluence measurements of the two samples are only consistent within 2σ(i.e.,p= 0.24) and the peak flux measurements are only just barely consistent at 2σ(i.e.,p= 0.05). Naively, since these are all high-zGRBs,...

2012
[6]

Low-zGRB Light Curves and Simulation Results 19 T able 5.GRBs in our high-zsample (i.e., z >3) GRB NamezT 90,true Fluence (sec) (cnts/det) 050319 3.2425 151.584 2.432 050502B 5.2 17.724 0.822 050505 4.2748 58.852 3.422 050814 5.3 142.852 3.06 050904 6.295 181.576 7.806 050908 3.3467 18.284 0.85 050922B 4.5 157.024 4.252 060116 6.6 104.832 3.579 060206 4.0...

2086
[2018]

During a slew, Swift does not trigger and becomes less sensitive to dim and short events

An alternative cause of the average peak flux change could be due an evolution in Swift’s observing strategy; over Swift’s lifetime, its average pointing interval has decreased from∼990 sec in 2005 to∼650 sec in 2025 and, consequently, 17 its time spent slewing between pointings has increased. During a slew, Swift does not trigger and becomes less sensiti...

2005
[2019]

– however, Figure 12 doesn’t show any average increase for high-zGRBs, or (ii) a decrease in the average redshift for Swift/BAT GRBs over time. In Figure 14Middle, we show the average redshift of Swift/BAT triggered GRBs over time and find a possible small decrease of the average redshift starting in∼2016, however the averages in 2022 and 2024 don’t suppo...

2016

[1] [1]

P., Abbott, R., Abbott, T

Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2017, ApJL, 848, L13, doi: 10.3847/2041-8213/aa920c Ahumada, T., Singer, L. P., Anand, S., et al. 2021, Nature Astronomy, 5, 917, doi: 10.1038/s41550-021-01428-7 Ajello, M., Arimoto, M., Axelsson, M., et al. 2019, ApJ, 878, 52, doi: 10.3847/1538-4357/ab1d4e Astropy Collaboration, Robitaille, T. P., Tollerud...

Pith/arXiv arXiv doi:10.48550/arxiv.2403.13126 2017

[2] [2]

Looking only at Swift/BAT GRBs that have measured redshifts, it seems this peak-flux increase may be more pronounced (see Fig. 12b). For GRBs with redshiftsz >3, this bias does not seem to be present, but the sample size is quite low, especially so after 2016 (see Fig. 12c). But even for GRBs without any redshifts, there seems to be a slight increase in t...

2016

[3] [3]

Naively, since these are all high-zGRBs, these measurements are drawn from the same underlying distribution and a KS-Test should findp≥0.32 to indicate this fact

We can see that duration measurements of the pre-2012 and post-2012 samples are consistent with one another to within 1σ(i.e.,p= 0.61), however, the fluence measurements of the two samples are only consistent within 2σ(i.e.,p= 0.24) and the peak flux measurements are only just barely consistent at 2σ(i.e.,p= 0.05). Naively, since these are all high-zGRBs,...

2012

[4] [6]

Low-zGRB Light Curves and Simulation Results 19 T able 5.GRBs in our high-zsample (i.e., z >3) GRB NamezT 90,true Fluence (sec) (cnts/det) 050319 3.2425 151.584 2.432 050502B 5.2 17.724 0.822 050505 4.2748 58.852 3.422 050814 5.3 142.852 3.06 050904 6.295 181.576 7.806 050908 3.3467 18.284 0.85 050922B 4.5 157.024 4.252 060116 6.6 104.832 3.579 060206 4.0...

2086

[5] [2018]

During a slew, Swift does not trigger and becomes less sensitive to dim and short events

An alternative cause of the average peak flux change could be due an evolution in Swift’s observing strategy; over Swift’s lifetime, its average pointing interval has decreased from∼990 sec in 2005 to∼650 sec in 2025 and, consequently, 17 its time spent slewing between pointings has increased. During a slew, Swift does not trigger and becomes less sensiti...

2005

[6] [2019]

– however, Figure 12 doesn’t show any average increase for high-zGRBs, or (ii) a decrease in the average redshift for Swift/BAT GRBs over time. In Figure 14Middle, we show the average redshift of Swift/BAT triggered GRBs over time and find a possible small decrease of the average redshift starting in∼2016, however the averages in 2022 and 2024 don’t suppo...

2016