User Tools

Site Tools


The Great Bruce GRB Info Library

Welcome to the famous Bruce GRB Info Library. Please comment and fix.

Special GRB List

080319b Prototype? Naked Eye Burst, probably optically brightest, best optical light curve, proposed two-component jet (Racusin+08
11025APrototype? Very long burst, optical correlates better with MeV than KeV (Guiriec+16)
130427ALargest E_ISO ever measured; 5.65e54 erg (1 keV-10MeV); Fermi and Swift - Ref: GCN 14576 E_iso; Swift GCN 14448
130603B - short, Bright burst
130427 Prototype? Well-opserved prompt optical, multiple reverse shocks (Vestrand+14)
160625BPrototype? 3 episodes with different spectra; Thermal, Band, PL (Zhang, B. B-, 2018NatAs...2...69Z)
GW170817 Short-GRB observed in LIGO/VIRGO, confirming for the first time NS inspiral model
190114C brightest LAT event ever; detected by MAGIC at > 300 GeV GCN 23701, intepreted as evidence for IC in GRB (Veres, P. 2019 Nature 575,459); z=0.425

Z>6 list of GRB

050904 6.295
060116 6.6
080913 6.695
090423 8.2
090429b 9.4

General GRB Observation Topics



Fundamentals - GRBs are beamed - The Inverse Compton Catastrophe

There is a lot to this. There is an old standard radio astro thing called the Inverse Compton Catastrophe - that is that if Tb > 1e12K, a radio source , given equipartition of particle and mag energy, will radiate all its energy away in the X-rays in days, because so many electrons will be upscattered, and radiate rapidly at high E.

The paper Readhead 94, ApJ 426,51 “Equipartiion brightness temp and the IC Catastrophe” discusses this, and shows that they don't really see this, but rather a maximum Tb at 1e11K (rest).

He also notes that there are brightness temp.s estimated by variability time - because the timescale for the e- to “poop out” is a function of Tb. This drops rapidly with Tb = 1e11K; intra-day variability implies Tb ~ 1e12k at high radio freq; ~ 1e13k at low. Tb measured by interferometer is limited by the resolution; vblty time I think is more sensitive or allows higher.

Conclusion: Very high Tb will scatter electrons to very high energies, and the vblty time scale will be very, very short, and requires constant energy injection. The way this is used is that if Tb> 1e12, and source is constant, it must be beamed.

I think there is another issue of pair production opacity, I did not see it in that paper.

keywords: Inverse Compton Catastrophe, brightness temperature limit, maximum brightness temperature, beaming.


X-ray Light Curves

'X-ray LC:'

Here is Zhang, B.+06 X-ray light curve diagram:


X-ray Spectral Slopes and Epeak

GRBs are often modeled as Band Functions - high-E and low-E power laws, but high-E is > 100 keV (for Fermi to observe) so we are mostly concerned with low-E slope. Nava, L et al. 2010 arXiv: 1004.1410 says:

Please note that Nava uses the term “index” to mean log slope, this is defined in her text; this is opposite the usual convention.

 Low-E index Please note that Nava uses “index” to mean log slope.

low-E Photon log slope is -0.92 for LGRB Epeak is 160 keV for Fermi/BATSE-like instrument (-1.57+/- 0.32 for BAT like instrument (peak of 79 keV; sakamoto+11) HE index is poorly constrained but is ~ -2.3 (Kaneko+2006)

low-E Photon log slope is -0.5 for SGRB, Epeak is 490 keV (-1.2+/- not given for BAT like instrument sakamoto+11)

where E_peak, discussed below, really means location of spectral break, technically peak in nu*fnu, which is flat for equal energy emitted per decade of frequency.

So if you are Swift, and fit a power law to the BAT results, you get a pretty different result from Fermi: photon index -1.57+/- 0.32, second BAT catalog, sakamoto+11: Sakamoto, T. et al., “The Second Swift BAT Gamma-Ray Burst Catalog”, ApJS, 195, 2 (2011),


(where EE = with extended emission). Why different results? Sakamoto said it best: “Therefore, we may conclude that “true” Eobs_peak has a single broad log-normal distribution. The difference of the Eobs_peak distributions among the GRB instruments is very likely due to an instrumental selection effect.” In other words, the distributions can be broad and only look peaked in a given instrument.

I note that Belobodorov 2013, has a paper , giving a distribution of 1-10 MeV for Epeak_Obs*(1+z), and a simple picture where this is changed by other factors further from the central engine to go to 30 MeV. Why this appears differernt from the 160-500 KeV Epeaks above, I don't know. He just quotes Goldstein 2012 (which is a Fermi Spectral catalog) and Kaneko 2006. Makes no sense.

Photospheric Emission

Joint LAT-GBM fits show a broadened BB, e.g. centered at ~290keV in 090902b, and which dominated the early part of emission in that burst, identified as emission originating in the Thompson photosphere of the jet. Ryde+ 2010,

Alternative Classifications to the Short vs. Long Classification Scheme:

'Spectral Lag Classification Scheme' In the GCN 12653, they (Barthelmy) went ahead and classified a burst not simply according to duration, but also according to spectral lag analysis. This could reference many papers, but Kann+11 Says Norris & Bonnell (2006) is the biggie, or a paper with classification in the title is, Gehrels, N., Norris, J., Barthelmy, et al., A New Gamma-Ray Burst Classification Scheme from GRB060614, Nature 444, 1044, 2006.

Norris & Bonnell (2006) propose that a strong indicator for a merger-population event is a negligible spectral lag between different energy bands (see also Zhang et al. 2006), i.e., these do not follow a lag–luminosity correlation (Norris et al. 2000).33

'Type I/Type II Classification Scheme' Type I = NOT massive star collapse (classical short) ; Type II = massive star collapse with SNe eventually coming out.

This is motivated by a simple classification: Has SN component, or not. You get this by very deep AG observations.

Kann+11 is a summary paper on afterglows… but somehow gets into re-classifying them based on AG properties.

spatial and luminosity distributions - Wanderman+Piran10 (keyword spam luminosity function, spatial distribution, number density)

<font size=1>'Note 1': On Classification Schemes: Quoting from Kann, D. +11, reviewing the Type I/Type II Classification Scheme: “a new classification scheme for GRBs (Zhang 2006; Gehrels et al. 2006; Zhang et al. 2007a). In analogy to Type Ia and Type II SNe, GRBs can be classified as Type I events (induced by the catastrophic destruction of a compact star or stars, no associated SN, can be found in all types of galaxies) and Type II events (induced by the destruction of a massive star, associated with an SN—most likely a broad-lined Type Ic SN—and found only in galaxies with high specific star formation). This definition is not based on a single observed quantity (such as the location in a T90-hardness diagram), and thus it is, in some cases, very difficult to place a GRB into the context of Type I or Type II. Still, we will adopt this ter- minology. Zhang et al. (2009) have extensively discussed the links between progenitors (collapsars versus mergers) and the expected observables, and have created a flowchart to help in the classification of Type I and Type II events (their Figure 8), which we shall employ. A second classification method which expands on Zhang et al. (2009) has been given by Lu ̈ et al. (2010), we will compare our classification results with this classification also. For an even more detailed approach toward GRB classification based on physical progenitor models, see Bloom et al. (2008).</font>



This paper: Ofek, E. O. +13 says that precursors to certain types of SNe are observable, and has a good example (z=0.035 I think). This suggests that GRB may have precursors.

keywords: GRB-SN connection; optical precursors; PTF palomar transient factory


Optical Light Curves:

Forget the picture of simple PL decay: Zaninoni+13 A&A 557, A12 (2013) “Gamma-ray burst optical light-curve zoo…” studies 68 GRB - very biased sample requiring z and good opt,x LC coverage, still:

  • simple pl decays are rare, ≤ 10/68 of sample (for bright well-covered sample)
  • 53 of 68 begin with rise or plateau 1e2 s to 1e4 s mode~ 1e3 s. ( ” “ )

Optical and X-ray LCs are different at early times in the majority of cases (Melandri et al. 2008b; Rykoff et al. 2009; Oates et al. 2009, 2011). In particular, Oates et al. (2009, 2011) noted that the optical LCs can decay or rise before 500 s after the trigger in the observer frame and do not show the steep decay as the X-ray LCs; after 2000 s the optical and X-ray LCs have similar slopes.

Afterglow temporal decay slope alpha is - 0.75 to -2.0; Liang+13, Liang, En-Wei et al. ApJ 774:13 2013

Optical SEDs:

GRB have ~-0.75 log slope power law for prompt optical says Rykoff, et al., 2009, ApJ 702,489;

Afterglows have more like -0.5 log slope Zaninoni+13 (

E(B-V) ~ 0.2 zaninoni+13

Prompt Optical

To understand the various spectral components and the theoretical slopes, please see the section on Emission Mechanism Theory.

GRB 110205A: This paper looks like this burst 110205a might be one of the best observed ever, rivaling the Naked Eye Burst 080319B.

Other Papers: This oddly short little paper gives a kind of list of “all” prompt optical, and may be useful:

BIggest collected study of prompt is Kopac+13, mostly ground-based.

Is prompt optical Reverse Shock (RS) Internal shock Synchrotron (ISS) or SSC?

130427 Vestrand+14 multiple RS
080319b (Racusin et al. no RS, multi-component jet with SSC)
061121 Page+07 could be ISS, no RS.…663.1125P
  • Opt>Spec

Initially it was proposed that MgII deep absorbers were more common in GRB than QSO; not any more:

  • Opt>Spec > redshift distribution - Sakamoto+11 SwiftCat2 has a median of z=1.8


SGRBs seem to come from old star populations, more ellipticals and halos. However, in a 2006 paper, Grindlay claimed that was due to narrower beams from SGRBs; binaries more common in disk populations will tend to have co-aligned spins and tehrefore have narrower beams.

'SGRB>X-ray Observations'

SGRB are 10% for Swift (sakamoto+11) 18+/-3% for Fermi GBM (Paciesas+12) and 24% for BATSE (Paciesas+12), based on T90 duration. Paciesas, W. S. et al., W. S. et al., 2012ApJS..199…18P, “The Fermi GBM Gamma-Ray Burst Catalog: The First Two Years”,

SHGRB are not just harder than LSGRB, their most dependable characteristics is nearly zero spectral lag; SHGRB often have low-level longer term (confusing called “extended” - meaning extended in time, not space) emission, which is SOFT BUT HAS NO LAG EITHER. Norris & Bonnell 2006,

Here, lag means peak or profile lag in 25-50 keV vs. 100-300 keV bands.

N&B06 make the interesting comment that relativistic beaming with different gammas in the two pop.s could account for the lag; the problem is that leg is not ruled out in softer emission. Sounds like a great opportunity for when SVOM and soft lobster optics instruments come on-line.

Alternative models: Virgili et al. 2011; Cui et al. 2010 (but see Leibler & Berger 2010) say that SGRB could not come from compact object mergers.

Bromberg, Omer +12 finds that merger GRBs dominate short durations, up to about 2 s for BATSE, but only to ~ 0.7 s for Swift. The physical process of LGRBs, an engine active for a finite time and a finite wait time for break-out, gives a plateau in duration distributions, which he shows is present, and emphasized when soft BATSE bursts are removed. So, Swift's plateau extends well down to 0.7 s. In an aside, this plateau effect will be present for anything preventing emission, including the optical distribution of rise times if there is absorbing dust that needs to be punched through (last sentences of paper).

'SGRB>Optical Observations'

Edo Berger Claims 25 identified short GRBs (2 in ULIRGS) as of 24 Sep 2012 21:11:58 (arXiv:1209.5423).


This paper claims that SGRB are from Kilonova. The evidence is maybe a little thin, but at least it's an interesting reference.

End of General Section

Bulk Lorentz Factor (BLF)

I usually quote Molinari et al. 2007, A&A 469,13 [] for his time to peak optical afterglow ⇐ > BLF; however, this paper (Chang, Zhe. et al. 2012, ApJ 759:129 discusses the BLF in more detail: and they use the delay time between 1 MeV and higher energy photons to “measure” the BLF, apparently only from LAT detected bursts. This is common in theory papers. Xiao-Hong Zhao et al. 2011 ApJ 726 89 doi:10.1088/0004-637X/726/2/89 reaches a similar conclusion to that above: the more zones you have, the lower your BLF.

Abdo et al. is the big group fermi etc. paper (actually a series) that gives a BLF whenever there is a good LAT detection; anyway for bursts with correlated low-high E you estimate BLF from variability time scale.

Look at my review of BLF measurements for more details.

GRB Extinction

Recent General Paper on GRB extinction: arXiv:1303.4743 Covino+13 - in this unbiased survey they say A_V_restframe 0.35 Mag is median Schady+2007 gives the figure for UVOT detected bursts: A_V_rest = 0.39 mag

It is quite interesting to note there at at least three papers on 2175 Å features in GRB spectra: arXiv:1205.0387 Zafar+14 and arXiv:0912.5435, arXiv:0810.2897

The original paper “GRBs are often extinguished A_V=1-5 mag” is Perley+09

Polarization Observations

  • Polarization> gamma/X

Yonetoku+11 Yonetoku, D., et al. 2012, ApJ 758, L1, Detection of polarization with IKAROS GAP.

  • Polarization> Optical

2012 Feb Still no detection of prompt optical, verified by the introduction of Yonetoku+11

Rather Detailed Optical Polarization Afterglow Phase measurements: Greiner, J. et al., 2003, Nature, 426, 157-159

  • Polarization> Theory

Lazzati 2006

Toma et al. 2009

GRB Relativistic Blast Waves

X, Opt, flares and re-brightening

This subject, the theory framework of GRB blast waves, mostly the afterglow, is covered in the Bruce GRB Blast Page.

On this page you will find the stages of evolution of the GRB blast wave.

There is also a very brief section mapping observations to theory. This is important in measurement of the BLF.

Correlations (possibly with z)

The Amati, Ghirlanda, and Yonetoku relations are the most famous. Many papers are out there, but 'many' papers show that there is huge uncertainty and risk in using these correlations to get the red shift. I can't believe that many people would take your work seriously, and reviewers would slaughter you.

Kocevski11:Kocevski, D. 2012, ApJ, 747, 146 - simulations of populations and accounting for instrument response shows how there is nothing physical in the correlations.

Shahmoradi&Nemiroff11: Shahmoradi, A. & Nemiroff, R. J. 2011, MNRAS, 411, 1843 or, same as above.

And again,, Colazzi+ (includes schaefer & Preece), accepted to ApJ.

'A universal scaling for short and long gamma-ray bursts': E_{X,iso}-E_{gamma,iso}-E_{pk} Bernardini, M. G.+12 , arXiv:1203.1060 (A universal scaling for short and long gamma-ray bursts: EX,iso - Eγ,iso - Epk, MNRAS, 425, 1199, M. G. Bernardini, R. Margutti, E. Zaninoni and G. Chincarini). The big result is a search for correlations of everything yields the best correlations need to include some other parameter of the afterglow.

The paper above is supposedly based on the exhaustive analysis of Margutti, R. 12+ arXiv:1203.1059, ( doi: 10.1093/mnras/sts066) but I see lots of data selection.

high-z GRB

Photo-z analysis suggests z~9.4 for GRB 090429B —, Cucchiara, A. et al. 2011, Astrophys. J. 736 (2011) 7

See Burrows, D. + 12 Because he gives the z>6 list of GRBs

Distribution of GRB into hi-z: Wanderman&Piran10 can tell you all about it!

* Hi-z> EOR_science> Epoch of Re-ionization and or end of cosmic dark ages.

Emission Lines from z=6.5 QSO - Role of AGN feedback in quenching star formation- important to EOR modeling - Maiolino+12 Using [CII] 158 um they got > 10 sigma detection of emission line at IRAM for z=6.5 QSO after 18.5 hours. The observation of weak winds implies star formation, while the observation her eof powerful winds indicates AGN-driven powerful winds which quenches star formation.

Using GRB080319 spectra to limit the neutral fraction at z= 6.7 Patel+10

Probably this shouldn't be here but I don't have much anywhere on absorption lines from Lya clouds at high-z. This section is not really relevant to GRB, it's just some keyword spam or something to lead to more useful info. At random I saw a paper on high-z Lya clouds in quasars, I might start with that: Garzilli+12 ArXiv #: 1202.3577

Check out the Bruce EOR Page which shouldn't go on a GRB page.

* Hi-z>Relativity

* relativity>time_dilation_signatures

Kocenski&Petrosian11 - Lack of time dilation in pulses is observational bias


Ellis&Mavromatos11. L-Invariance and CTA Cherenkov Telescope Array.

HECR_associated_with_GRB - Lots of good stuff about cerenkov teles which slew to GRBs, how they can observe at ~ 1e2 GeV (I think they mean photons). They are talking about new generations of Imaging Cerenkov Atmospheric telescopes.


* Obs> other_events>tidal_disruption_events

Krolik&Piran11 - this is about Swift J2058

* Obs> other_events> XRF

What is the difference between an X-ray Flash and a GRB?

All I know if that they are softer and less luminous in X-gamma (be careful they can be the same explosion energy if you total all bands), but still can be detected to high-z (e.g. 2). Here is a crap popular page that says they leave magnetars behind, but it does not sound serious work to me. There is a nature article referenced here that says they are radio-bright

Instrumentation & Observations

* X-ray_observations - includes background and band selection

* Swift — See the Swift_instrument page, about the BAT, XRT, and UVOT.

* BATSE— Here is the effective area (efficiency) curve: batse_response.png “The LAD detector is a disk of NaI scintillation crystal 20 inches in diameter and one-half inch thick, mounted on a three-quarters inch layer of quartz. The large diameter-to-thickness ratio of the scintillation crystal produces a detector response similar to that of a cosine function at low energies where the crystal is opaque to incident radiation.”—

* Fermi -The main instruments are the

'GBM', which is like a small BATSE,

Low-E NaI - covers from ~ 8 keV to 1 MeV

High-E BGO - covers ~150 keV to ~30 MeV

and the

'LAT', which is a monster Si tracker which gets all the famous high-energy photons, I think up to ~ 1e2 GeV

20 MeV - 300 GeV

Active Shielding/ Hybrid Detector Instruments - Below I give instruments that are mostly low-E, semiconductor detectors. One big differerence between HE and LE instruments is the lack of active shielding. For more info on active shielding or a kind of hybrid detector, see

'Semiconductor Instruments'


* AGILE: Longo+11,

SuperAgile - Super-AGILE (SA), the ultra-compact and light hard-X-ray imager of AGILE is a coded-mask system made of a Silicon detector plane and a thin Tungsten mask positioned 14 cm above it. The detection cabability of SA includes: (1) photon-by-photon transmission and imaging of sources in the energy range 18-60 keV, with a large field-of-view (FOV 1 sr); (2) an angular resolution of 6 arcmin; (3) a good sensitivity (15 mCrab between 18-60 keV for 50 ksec integration).

* Janus


<div id=“Burrows+12”></div>Optimization for high-z GRB: (Lobster Optics vs. standard coded mask) Burrows+12 or Burrows, D. N. et al., “Optimizing the Search for High-z GRBs: The JANUS X-ray Coded Aperture Telescope” Basically, he finds that wide area of coded mask cameras trumps lobster optics (much less sky area/Vol/mass and wider energy). He does make the prediction that Epeak ~ 200 keV/(1+z) so HZGRB peak at ~ 30 keV or less. He uses z=6 as the fiducial z for high z because that's supposed to be the end of the EOR.


XCAT - Falcone+10 - Proc. of SPIE Vol. 7732 77324F-7, is the X-ray coded mask camera on Janus. The detectors are Hybrid CMOS Si, produced by teledyne and read out with the Teledyne Sidecar ASIC! These are sensitive ~ 0.5-20 keV, i.e. pretty soft!

* Astro-H - DSSSDs and double sided CdTe strips!

* MAXI - 2-30 keV Gas Slit Camera. MAXI is really interesting because it is on the ISS and you can see the background measurements in this paper - Sugizaki, M. et al. 2011 (In-Orbit Performance of MAXI Gas Slit Camera (GSC) on ISS) -

* EXIST - Josh Grindlays super huge mission, now killed, to high GRBs z> 7 by going down to 5 keV w/CZT detectors.

Grindlay+2010 2010SPIE.7732E..59G ; LOTS of details given in protoEXIST baloon flight paper

What I don't understand is that in his APS presentation Grindlay claimed response to 600 keV; he gives CZT as 2 cm X 2 cm (0.6 mm pixels) X 0.5 cm deep. What response do you get with 0.5 cm deep? For comparison, SVOM xtals are 1 mm thick, so maybe this is nonsense???

Predicting GRB Rates

See this page: grbratepredict

Explosion Theory

(Theory papers defined as ones that have little or no data analysis, mostly simulations; Generally speaking, look first in the subjects above.)

Hyperaccretion - It's the opacity, stupid. Begelman 12, arxiv 1203.1628 - he says its all about opacity, not ang. Mom. Thirdly, if he says there is super-Eddington accretion, he must have some ideas about how to get rid of ang. mom.

Angular momentum and rotation in LGRB. Dessart+12 ArXiv #: 1203.1926 - This is a meaty paper, which says rotation and angular momentum are critical (not sure it's greatest paper however. They make the point again that these things are hard to get to explode.

Emission Mechanism Theory

Emission Mechanism Theory

General Gist: There are spectra, models, and then Emission Mechanism.

Spectra: Mostly GRB have BAND function spectra, which can only sort-of be explained by a synchrotron emission mechanism. However, optical never quite fits these predictions. ELIMINATED: Synch. Self Compton (SSC) is eliminated in all observations, as Fermi-LAT has never seen the high energy upscattered part of that emission.

Synchrotron Emission Mechanism Theory (Encoded information in self-absorption frequency)

Shen & Zhang 09 paper 2009, MNRAS 398, 1936 shows that Synchrotron can have the same appearance at gamma, X, but very different manifestations at optical. Here are the four cases:

Inverse Compton Emission

This paper, on IC in prompt GRB emission, , Yue Zhang+19 ApJ, 877,89 abstract ends with “Therefore the ICS component is more likely to be detected for GRBs that have a hard low-energy photon spectrum.”

The paper says that big BLF ⇔ adiabatic expansion dominant cooling of e-s, ⇔ hard LE photon spectrum ⇔ Observable IC.

NOTE: IC NOT YET OBSERVED IN PROMPT - 190114C MAGIC ~ 1 TeV gammas intepreted as evidence for IC in GRB, but in AFTERGLOW (Veres, P. 2019 Nature 575,459). The MAGIC light curve is clearly power law decay, not prompt.

What are the prospects for observing prompt IC? What are the response times of Hess and Magic and Hawc? Apparently, Magic, HESS responds to GCNs, as I believe do all atmospheric cerenkov telescopes. ONLY HAWC has a wide FOV and is not steered, making it ideal for prompt, but it has yet to detect any GRBs.

HAWC - water cherenkov, not steered, so can measure prompt! FOV ~ 15% of sky ( No GRB detections yet.

MAGIC - steered UV/opt ???9 sq deg FOV - can't tell from website.

HESS - Stereoscopic UV/opt in Namibia, steered, one set of teles has 3.2 deg, one 5.0 deg on sky.

Other components, multiple components, etc.

Guiriec saw an MEV component in a really long burst that correlated with UVOT emission much better than BAT emission. I think I see the same thing in the Zhang+18 GRB160625B data, and they don't even mention that! - weird - all they say is that there is a ~ 3 sec delay between GBM and optical, similar to 080319b.

Some GRB have thermal spectra at some point (e.g. Zhang+18 GRB160625B), which should be from an optically thick photosphere.

Models: Forward shock (fast variable, bright gamma-rays) , reverse shock (e.g. Vestrand+14), and even external shock afterglows, can all be from synchrotron mechanism.

Cosmology Theory with GRB

Lu+12 arXiv:1203.4907 explores Generalized holographic DE models using many data sets including Hi-Z GRB.

Correlation Studies

Correlations (possibly with z)

The Amati, Ghirlanda, and Yonetoku relations are the most famous. Many papers are out there, but many papers show that there is huge uncertainty and risk in using these correlations to get the red shift.

Kocevski11:Kocevski, D. 2012, ApJ, 747, 146 - simulations of populations and accounting for instrument response shows how there is nothing physical in the correlations.

Shahmoradi&Nemiroff11: Shahmoradi, A. & Nemiroff, R. J. 2011, MNRAS, 411, 1843 or, same as above.

And again,, Colazzi+ (includes schaefer & Preece), accepted to ApJ.

A universal scaling for short and long gamma-ray bursts: E_{X,iso}-E_{gamma,iso}-E_{pk} Bernardini, M. G.+12 , arXiv:1203.1060 (A universal scaling for short and long gamma-ray bursts: EX,iso - Eγ,iso - Epk, MNRAS, 425, 1199, M. G. Bernardini, R. Margutti, E. Zaninoni and G. Chincarini). The big result is a search for correlations of everything yields the best correlations need to include some other parameter of the afterglow.

The paper above is supposedly based on the exhaustive analysis of Margutti, R. 12+ arXiv:1203.1059, ( doi: 10.1093/mnras/sts066) but I see lots of possible data selection.

public/bggrbinfo.txt · Last modified: 2020/04/21 01:16 by bruce