NASA Technical Reports Server (NTRS) 20120010113: Broadband Study of GRB 091127: A Sub-Energetic Burst at Higher Redshift?

5.7 MB·English

by  NASA Technical Reports Server (NTRS)

DRAIT '.'ERSJON JANUARY 23, 2012
Prep~int typeset using ~TgX style emulateapj v. 04/21/05
BROADBAND STUDY OF GRB 091127: A SUB-ENERGETIC BURST AT HIGHER REDSHIFT?
E. TROJAt•2, T. SAKAMOT02, C. GUtDORZI3,"" J. P. NORRlSs, A. PANAITESCU6:! S. KOBAYASH]<i, N. OMODEl7, J. C.
BROWN2, D. N. BURROWSs, P. A. EVANS9, N. GEHRELS2, F. E. MARSHALL , N. UAWSON\ A. MELANDRltO.4
C. G. MUNDELL4, S. R. OATES II , V. PAL'SHINL2, R. D. PREECE13, J. L. RACUSIN2. 1. A. STEELE", N. R. TANVlR9,
V. VASILEJOUl4, C. WILSON-HODGE 5, K, YAMAOKAI6
Draft version January 29, f!012
ABSTRACT
GRB 091127 is a bright gamma-ray burst (GRB) detected by Swift at a redshift z=0.49 and as
sociated with SN 2009nz. We present the broadband analysis of the GRB prompt and afterglow
emission and study its high-energy properties in the context of the GRB/SN association. While the
high luminosity of the prompt emission and standard afterglow behavior are typical of cosmological
long GRBs, its low energy release (E,<3x1049 erg), soft spectrum and unusual spectral lag connect
this GRB to the class of suh-energetic bursts. We discuss the suppression of high-energy emission in
this burst, and investigate vrhether this behavior could be connected vrith the sub-energetic nature of
the explosion.
Subject headings: gamma-ray bursts: individual (GRB 091127)
1. INTRODUCTION 2003), GRBs with spectroscopically confirmed SNe show
a peculiar behavior, both in their prompt and after
It is well established that (most) long duration GRBs
glow emission phases (Kaneko et a!. 2007; Starling et a!.
are linked to the gravitational collapse of massive stars
2011). These bursts are characterized by a relatively
(Woosley & Bloom 2006). Such a connection is sup
ported by several lines of evidence (Hjorth & Bloom softer spectrum (Epk ~120 keV) , and a lower energy out
2011, 2..nd references therein). In a few remarkable cases put (E"i,o~1048-1050erg) than standard GRBs. They
the spectroscopic identification of a broad line Type Ie do not strictly follow the lag-luminosity relation (Norris
SN, co-spatial and coeval with the GRB, provided a di 2002), whereas they generally agree with the Amati re
,..... rect proof of the physical association between the two lation (Amati et aI. 2007), but with GRB 980425 be
phenomena, ing a notable outlier. Suh-energetic nearby bursts tend
> With the exception of GRB 030329, whose properties to show a faint afterglow emission, both in X-rays and
- are roughly similar to typical long GRBs (Berger et al. in the optical band. Late time radio monitoring of
00 their afterglows showed evidence of a quasi-spherical and
..q-. 1 NASA Postdoctoral Program Fellow only mildly relativistic (r", 2) outflow (Soderberg et a!.
2 NASA, Goddard Space Flight Center, Greenbelt, MD 20771, 2006), very different from the highly relativistic and col
USA
limated jets observed in long GRBs (Bloom et a!. 2003;
3 Ph~'sics Department, University of Ferrara, via Saragat I, 1-
l\folinari et a!. 2007; Cenko et a!. 2010). For these rea
44122, Ferrara, Italy
4 Astrophysics Research Institute, Liverpool John MoorM Uni sons it has been speculated that suh-energetic e,-ents
versity, TwelYe Quays House, Egerton Wharf, CH41 lLD, Birken belong to an intrinsically distinct population of bursts
head,UK
IS PhY3ics Department, Boise State University, 1910 University which dominate the !ocal (z ~ 0.5) rate of observed
events (Liang et a1. 2007; Chapman et aI. 2007).
Drive, Boise, ID 83725, USA
6 Space Science and Applications, MS 0466, Los Alamos Na Whereas the case for spectroscopically confirmed SNe
tional Laboratory, Los Alamos, NM 87545, USA remains confined to nearby GRBs, at higher redshifts
7 W. W. Hansen Experimental Physics Laborator::-, KaYli In (0.3 < z < 1) the emergence of the associated SN is pin
stitute =or Particle Astrophysics and Cosmology, Department of
Physics and SLAC National _\ccelerator Laboratory, Stanford Uni pointed by a late-time optical rebrightening or c;bump" in
versity, Stanford, CA 94305, USA the afterglow light curves (Bloom et aI. 1999; Zeh et a1.
B Department of Astronomy and _\strophysics, Pennsylvania 2004; Tanvir et a1. 2010). Though alternative explana
State University, 525 Davey L~b, University Ps.rk, fA 16802, USA tions for such a feature are plausible (Esin & Blandford
9 X-rc:.y and Observational Astronomy Gro.uP, Department of
P~sics and Astronomy, University of Leicester, LEI 7RH, UK 2000; Waxman & Draine 2000), a spectroscopic analy
o INAF-OAB, via Bianchi 46, 1-23807 Mente (LC), Italy sis of some of these SN bumps supports their similar
11 Mullard Space Science Laboratory, University College Lon ity with bright Type Ic SNe (e.g. Della Valle et a1. 2006;
don, Holrnbury St. Mary, Dorking Surrey, RH5 6NT, UK Sparre et a1. 2011). This is the case of GRB 091127,
12 Ioffe Physico-Technical Institute, Laboratory for Experimen
detected by the Swift satellite (Gehrels et a1. 2004) at
tal _\sttophysics, 26 Polytekhnicheskaya, St Petersburg 194021,
Russian Federation a redshift of z =0.49, and associated with SN2009nz.
13 University of Alaba.ma in Huntsville, NSSTC, 320 Sparkman Cobb et aI. (2010) identified in the GRB afterglow a
Drive, Huntsville, AL 35805, USA
late-time optical rebrightening, peaking at a magnitude
14 Laboratoire Univers et Part.icules de ~rontpellier, Uni-.-ersite
Montpellier 2, and CNRSjIN2P3, h1ontpellier, France of 1= 22.3±0.2 mag at ~22 d after the burst, and at
15 Institute of Astro and Particle Ph)'Hics, University Innsbruck, tributed it to the SN light. The photometric prop
Technikerstrasse 25, 6176 Innsbruck, Austria erties of SN2009nz resemble SNl998bw (Galama et a1.
16 Department of Physics and 11athematics, Aoyama Gakuin
1998), though displaying a faster temporal evolution and
University, 5-10-1 Fuchinobe, Chuo-ku, Sagamihara, Kanagawa.
a slightly dimmer peak magnitude. More recently, the
252-5258
2 E. Troja et al. .
spectroscopic analysis presented bv Berger et al. (2011) subtracted in the usual fashion and flat fielded using a
uncovered the typical undulations of broad line Type Ie stack of twilight exposures. Standard aperture photom
SNe associated with nearby GRBs, thus confirming the etry was carried out using two local reference stars, and
SN origin of the photometric bump. Berger et al. (2011) calibrated by comparison with R band frames of the same
concluded that the explosion properties of SN2009nz field.
(EK",2xl051erg, M,;-1.4M0 , and MNI",,0.35M0) are Due to an Earth limb constraint, Swift did not im
remarkably similar to SN2006aj (Pian et al. 2006), asso mediately slew to the burst location and follow-up ob
ciated with GRB 060218. GRB 091127 therefore repre servations with its two narrow field instruments, the X
sents one of the best cases linking long GRBs and SNe Ray Telescope (XRT; Burrows et a1. 2005) and the Ultra,
at red,hifts z>0.3. Violet Optical Telescope (UVOT; Rorning et al. 2005),
While previous works mainly focused on the proper began 53 min after the trigger. As the X-ray afterglow
ties of SN2009nz and its environment (Cobb et al. 2010; was still bright (~1O ctss-'), XRT started collecting
Vergani et al. 2011), in this paper we present a broad data in Windowed Timing (WT) mode, and automati
band analysis of the GRB prompt and afterglow emission cally switched to Photon Counting (PC) mode when the
and study the high-energy properties of the explosion in source decreased to ':::;2 etas- I. Follow-up observations
the co~text of GRBjSN associations. Being a bright and monitored the X-ray afterglow for 36 d for a total net ex
relatively nearby burst, GRB 091127 has a rich multi posure of 760 sin WT mode and 470 ks in PC mode. The
wavelength coverage up to very late times, which allows optical afterglow was detected by UVOT in the White,
us to study in detail its spectral and temporal evolution V, u, uvwl, and uvm2 filters at a position consistent
(see also Filgas et 1'1. 2011) and compare it to other well with the LT localization. The detection in the UV fil
known cases of GRBsjSNe. ters is consistent with the low redshift z=0.49 of this
The paper is organized as follows: our observations burst. SwiftjXFIT and UVOT data were reduced using
are detailed in § 2: In § 3 we present a multi-wavelength the HEASOFTlT (v6.11) and Swift software (v3.S) tools
timing and spectral analysis of both the prompt and and latest calibration products. We refer the reader to
the afterglow emission; our results are presented in § 4 Evans et al. (2007) for further details on the XRT data
and discussed in § 5. Finally, in § 6 we summarize our reduction and analysis. The UVOT photometry was
findings and conclusions. Throughout the paper, times done following the methods described in Breeveld et al.
are gh'en relative to the Swift trigger time To, t=T-To, (2010) with adjustments to compensate for the contam
and the convention fv,t ex: lI-(3t-a has been followed ination of a nearby star.
l
where the energy index f3 is related to the photon index In order to monitor the late time X-ray afterglow, two
r = f3 -l-1. The phenomenology of the burst is presented Target of Opportunity observations were performed by
in the observer's time frame. Unless otherwise stated, the Chandra X-Ray Observatory at t=98 d for a total
all the quoted errors are given at 90% confidence level exposure of 38 ks and t=IS8 d for a total exposure of
for one interesting parameter (Lampton et al. 1976). SO ks. Chandra data were reduced using version 4.2 of
the CIAO software. Source events were extracted from
a 2 pixel radius region around the GRB position, while
2. OBSERVATIONS AND DATA REDUCTION
the background was estimated from a source-free area
GRB 091127 triggered the Swift Burst Alert Tele using a 20 pixel radius region.
scope (BAT; Barthelmy et al. 2005) at 23:25:45 UT on
2009 November 27 (Troja et al. 2009). It was also ob
served by Konus-Wind, Suzaku Wide-band All-sky Mon 3. DATA ANALYSIS
itor (WAM), and the Fermi Gamma-Ray Burst Monitor
3.1. Gamma-ray data
(GBM). The burst was within the field of view of the
Fermi Large Area Telescope (LAT; Atwood et al. 2009), 3.1.1. Temporal analysis
at an angle of 25' from the boresight.
Figure 1 presents the prompt emission light curves
The 2-m Liverpool Telescope (LT) responded robot
with a 128 rns time resolution and in four different en
ically to the Swift alert and began observing at
ergy bands. The burst duration, defined as the inter
23:28:06 UT, 141 s after the BAT trigger. The
val containing 90% of the total observed fiuence, is Too
detection mode of the automatic LT GRB pipeline
(15-350 keV)=7.1±0.2 s. The burst temporal profile is
(Guidorzi et 1'1. 2006) identified a bright optical af
characterized by two main peaks, at t.....,Q sand t.....,1.1 s,
terglow (r' = 15.4 mag) at " = 02h26wl9'!89, 0 = respectively. They are clearly detected up to -600 keY
- IS'57'OS'!6 (J2000) (uncertainty of 0:'5; Smith et a1. and display a soft-to-hard spectral evolution. A period
2009). Observations were obtained with r'i' z' filters un of faint, spectrally soft emission lasting -S s, follows. On
til 2.3 hours post burst. The afterglow was monitored top of it a third peak at t~7 s is visible at energies below
with both the Faulkes Telescope South (FTS) and LT up 50 keY.
to 6 days post-burst within the BV Rr'i' filters. Magni Spectral lags were calculated by cross-correlating the
tudes of field stars in BV R were calibrated using Landolt light curves in the standard BAT channels: 1 (15-
standard stars (Landolt 1992) obtained during following 25 keY), 2 (25-50 keY), 3 (50-100 keV), 4 (100-350 keY).
photonetric nights. SDSS r'i' z' magnitudes of the same In order to increase the signal-to-noise ib the higher en
field stars were obtained using the transformations by ergy channels, the analysis was performed on non mask
Jordi et aI. (2006). Early time observations were also weighted lightcurves, each '\\'ith a 8 IDS time resolu-
obtained using SkycamZ, mounted on the LT tube. Ob tI· on. W e d en.v e d T31 = 2 . 2+- 2121..83 rns an d 7"42 = - 9 . 2+- 86..52 rns,
servations are filter-less (white light) to maximize the
throughput of the optics. The data were dark and bias 17 http://heasarc.gsfc.nasa..gov/docs/softwa:ce/lheasoft/
Broadband study of GRB 091127 3
fuse emission due to the interaction of cosmic rays with
4 BAT 15-50 ke the gas and the interstellar radiation field.
No significant excess above background was found.
,. 2 Following the procedure described in Abdo et al. (2009)
and by fixing the photon index to 2.25, we derived a 95%
upper limit of 2.8xlO-8ergcm-2s-1 in the 100 MeV-
,.
~Q) 0 1 GeV energy range and of 1.6xlO-8ergcm-2s-1 in
"0
the [. GeV-1O GeV energy range during the prompt emis
2 BAT 50-150 ke
'" sion interval (-0.3s<t<8.2s).
~§
1
0 3.1.3. Spectral analysis
U
o We performed a time-averaged and a time-resolved
spectral analysis, selecting the time intervals in corre
2000 150-300 keY spondence of the three main pulses as shown in Figure 1
(top panel). The spectral fits were performed in the 15-
150 keY energy band for BAT, 20 keV-lO MeV for Konus
1000
Wind, and 120 keV-3 MeV for Suzairur WAM. Following
Sa.kamoto et al. (2010) we added a 5% systematic error
o
in the WA11 spectra below 400 ke V. The intercalibra
§ tion between BAT, Konus-Wind and SuzakvrWAM was
600 BGO 300-600 keY
extensively studied by Sakamoto et a1. (2010), show
Uo 400 ing an overall agreement in the effective area correction
«20%) between the three instruments. GBM data were
200
fit in the 8-860 ke V band for the N aI detectors, and in
o
the 200 keV-40 MeV for the BGO detector. Given the
brightness of this burst we added a 5% systematic error
o
-5 5 10 to the GBM data, needed to improve the fit acceptance
Time since BAT trigger (s) of the time-averaged analysis. A cross-calibra.tion study
has not been performed Vlith the Fermi data yet. Pre
FIG. 1.-Swift/BAT (top two panels) and Fenni/GBM (bottom vious works (e.g. Page et 001. 2009) report a typical effec
wtwitoh paa nl2e8ls )m sb abcikngnrinogu.n dT-hsueb gtrreayc taerde alsig lhatb eclluerdv 8~8 o4f, bGaRnEd c0 9in1 1th27e tive area correction factor of ~ 1.23 compared to a value
top panel show the three intervals selected for the time-resoh-ed of unity for BAT, and in our analysis we found consistent
spectral analysis. Error bars are 1 u. values.
The best fit spectral parameters were estimated us
where the quoted uncertainties (at a r (j confidence ing the maximum likelihood method and, when neces
level) were evaluated by simulations. Lag analysis re sary, by applying different statistics to the data. BAT
veals a significant difference between the two main 'Y-ray mask-weighted spectra have Gaussian distributed uncer
peaks (a and b in Fig. 1). The former shows positive tainties, and they require the X2 statistics to be applied.
lags, 'T31 =36!~: rns and 742 =16~~~ rns, while the lat LAT spectra are instead characterized by low counts, and
ter has negligible or negative lags, T31 =-2~i~ IllS and they can only be modeled using the Poisson distribution.
In order to properly account for the Poissonian nature
742 =-14~ims.
of the source counts and for the Gaussian uncertainties
3.1.2. Search for high-energy "(-ray emission associated to the LAT background model (Abdo et al.
2009), we used the profile likelihood statistic as imple
The Fermi/LAT data were searched for emission dur
mented in the option PGSTAT of XSPEC (Arnaud et al.
ing the prompt ,)-ra.y phase and over longer timescales 2011). Table 1 reports the spectral fit results for the
(up to 10 ks). The searches were performed by means
time-averaged analysis. Different spectral models, usu
of an unbinned likelihood analysis (Abdo et al. 2009).
ally adopted to describe the GRB prompt emission spec
We used the Pass7V6 Transient class events with a re
trum, were fit to the data: a power-law (PL), a power-law
constructed energy above 100 l\leV. We selected events with a high-energy cut-off (CPL; F(E) ()( E"'e-E/E, •• ), a
within 12 degrees around the best burst position (see §2),
Band model (Band et al. 1993), and a Band model with
and applied a cut on zenith angle at 105° in order to
a high energy cut-off (Band+Cut). We also included the
limit the contamination from the bright Earth's limb.
log-parabolic function (LOGP; F(E) ()( EQ+':'logE) sug
For the Transient data class the dominant background
gested by Massaro et 001. (2010). The last column of Ta
component is the isotropic background due to residual
ble 1 reports the fit statistics (STAT) and degrees of free
charged particles misclassified as "(-rays. We modeled
it by using the tool developed by the LAT collabora dom (d.o.f.) for each model. In general STAT=X2, when
LAT data were included in the fit STAT=X2+PGSTAT.
tion that can predict the hadronic cosmic ray and ,,(-ray
Additional models, not reported in Table I, were
compo:J.ents of the background with an accuracy of ~W-
tested. A single-temperature black body plus a power
15% (Abdo et a1. 2009). We also addecl the template
gaL2yearp7v6_vO. fitslS describing the Galactic dif- law yields a poor fit (STAT /d.o.f=905/574), the addi
tion of a high-energy cut-off significantly improves the fit
18 Available at the Fecmi Science Suppoct Center web ,ite (STAT /d.o.f=675/573), but the model is not statistically
http://fermLgsfc.nasa.gov/ssc/data/access/lat/Bs.ckgroWldModels.html preferred to the standard Band function with a high en-
4 E. Troja et al.
TABLE 1
SPECrRAL FIT REsCLTS OF THE TIME-AVERAGED ANALYSIS
Det-ector Model Q /3 E.k (keY) E.ou. (keV) STATld.o.f
BAT PL 2. 17±O.07 52/57 (0.91)
KW PL 2. 19±0.04 74/59 (1.26)
KW CPL 2.oo±O.10 510:!:1~ 57/58 (0.98)
WAM PL 2.35:!:g:~: 23/25 (0.90)
WAM CPL 1.9±O.5 >400 20/24 (0.85)
GBM LOGP 0.73±0.14 0.37±0.04 465/396 (1.18)
GBM Band 1.20±O.16 2.23±0.04 39±5 457/395 (1.16)
GBM Band+Cut O.3:::~:~ 1.94±O.OB 25:!:~6 500:!:~gg 448/394 (1.14)
GBM+LAT LOGP 0.73±0.14 0.37±0.04 468/398 (1.18)
GBM+LAT Band 1.34±0.16 2.32±0.06 45±5 485/397 (1.21)
GBM+LAT Band+Cut O.6+o.~ 1.96:g:~~ 26:!:~5 530:!:1~ 450/396 (1.14)
JOINT LOGP 081 ;:O:J3 0.35±0.04 632/544 (1.16)
. - 0.10
JOINT Band 1.37±0.12 2.31±O.05 45±4 640/543 (1.18)
JOINT Band+Cut 1·06+~·~5 2.0 7+0g. Q128 36+~2 800+8ag0!0:! 602/542 (1.11)
TABLE 2
SPFCI'RAL FIT REsULTS OF THE TIME-RESOLVED ANALYSIS
Detector Model -Q -/3 Epk (keV) E.ou. (keV) STAT/d.o.f
Time interval a.: from To-O.3s to To+o.7s
BAT PL 1.91±O.10 59/57 (1.03)
WAh! PL 2.42!g:~~ 39/34 (1.15)
WAM CPL 1.89±0.5 600+2000 33/33 (1.00)
-300
GB~I LOGP <0.017 0.54±0.02 313/270 (1.16)
GBM Band 0.54±0.16 2.27±0.07 56±5 257/269 (0.95)
GBh! Band+Cut O.4:!:g:~8 1.97±0.17 54±6 600:!:~gg 247/268 (0.92)
GBM+LAT LOGP <0.019 0.55±0.02 314/272 (1.15)
GBM+LAT Band 0.60±0.1S5 2.32±0.06 59±5 266/271 (0.98)
CBM+LAT Band+Cut O.4:!:g:i 1.97±0.17 54±S 600+900 248/270 (0.92)
-200
JOINT LOGP <0.021 0.54±0.02 413/365 (1.13)
JOINT Band 0.63±O·13 2.34±0.06 59±5 369/364 (1.01)
JOINT Ba.nd+Cut 0.41+:2:28 2.02±0.11 53±5 700+600 344/363 (0.95)
-300
Time interval b: from To+O.8s to To+l.7s
BAT PL 1.78±0.12 52/57 (0.92)
WAM PL 2.38±0.11 34/34 (1.00)
WAH CPL 1.8:!:g:! lOOO:!::ggo 28/33 (0.87)
GBM LOGP 0.35±0.16 0.38±O.05 263/270 (Om)
GBM Band 1.22+0.08 2.23+0.2 14O±30 257/269 (0.95)
GBM Band+Cut I . 22;-:08.:113~ 2 . 13;-:80:.11 33 14O±30 >900 257/268 (0.96)
GBM+LAT LOGP 0.33±O.15 O.37±0.05 263/272 (0.97)
GBM+LAT Band 1.31±0.06 2.6!g:~ 170±30 2.56/271 (0.94)
GBM+LAT Band+Cut 1.30±O.07 2.52±0.17 170±30 >700 252/270 (0.93)
JOINT LOGP O.292tt~ 0.38±0.04 366/364 (1.00)
2.51:g:i;
JOINT Band 1.32±0.06 170!~g 361/363 (0.99)
JOINT Band+Cut 1.29+g·~~ 2.34±0.12 160+~ >1000 356/362 (0.98)
ergy cut-off (STAT/d.o.f=652/573). A multicolor black 3.2. X-my data
body (Ryde et al. 2010) gives similar results.
The XRT light curve is well described
Table 2 reports the results of the time-resolved spectral
(X2/d.o.f.=376/364) by a power law decay with
analysis for both intervals a and b. As found for the time
slope 0b=1.03±0.04 steepening to 02=1.S5±0.03 at
integr~ted spectrum, alternative models do not provide tbk=32~6 ks. The two Char.dm detections lie slightly
an improvement in the fit statistics and are not reported
above the extrapolation of this model, but are consistent
in the table.
with it within 3 u. This constrains the time of anr
The apectrum of the third peak (intervai c in Fig. I)
late-time jet-break in the X-ray light curve to t:2;115 d.
is well described by a power law of photon index
This time was determined by forcing in the fit an
r T=2.78±0.18. The average observed flux during this
BA additional break with :'0=1, and by varying the break
interval is 8~U)(1O-7 ergscm-2 S-I in the IS-SOkeV time until a Ll.X2=2.706 was reached.
band.
During our observations a slight soft-to-hard spec
tral evolution is visible over the first few hours. We
performed time-resolved spectral fits on seven consec
utive time intervals, selected according to the light
Broadband study of GRB 091127 5
§~~i~
10
o. ~h~.5 >:
10 ~ t:J•.0. ,VR8. +++1O2...S51 9 ~I "-"-"-
i -. '- _... 7j';-'1-2.S.5 10-1 ~'" 0.1 "-
qQ~~= . "'.>.":; "-"-"-"-
'" 15
0.01
~
20
10-3
1000 104 105 HI' 10' om
Time since trigger [sJ 0.1
Energy [keV[
FIG. 2.-X-ray and optical afterglow light curVE'S of GRB 091127
with the best fit models overplotted (solid and dashed lines respec FIG. 3.- Afterglow spectral energy distributions at 6 ks and
tively). At late times (t>lO d) the optical emission is dominated by 55 ks. The best fit model (solid line) and the same model corrected
the underlying host galaxy. Error bars are 1 (J', Optical magnitudes for extinction and absorption effects (dashed line) are shown.
are not corrected for Galactic extinction.
curve phases and to have'" 1000 net counts each. The are: ",=0.58±0.12, tbk,1=330:':}go s, "2=0.27±0.01,
X-ray spectra were modeled with an absorbed power tbk,2=4.1:':g·7 ks, "3=0.55±0.1O, tbk,3=28:':g ks,
law. We derived an intrinsic NH=9~~xl020 cm-2 at "4=1.34±0.04. Contamination from the SN-bump
z=0.49, in excess of the Galactic value of 2.8x 1020 cm-2 and the host galax)' light, not detected in the early-time
(Kalberla et al. 2005). The resulting photon indices LT exposures, may explain the shallower temporal
fx, ranging from 2.02±0.1O to 1.82±0.09, are consistent index at late times. By including in the fit a constant
within the uncertainties, however a systematic trend of component with magnitude I=22.54±0.1O to account
a slowly decreasing r x is evident. The time-averaged for the host emission and a SN-like bump, based on the
photon index is fx=1.88±0.08. observation of Cobb et al. (2010), the afterglow slope
Because of the low number of events in the Chandra steepens to "4=1.64±0.06.
spectrum (67 net counts) we used the Cash statistics
(Cash 1979) and fit it with an absorbed power law by 3.4. Spectral energy distribution
fixing the absorption components to the values quoted
An optical-to-X-ray spectral energy distribution (SED)
above. The resulting photon index is f x=1.6±0.3,
was produced at two different times, 6 ks and 55 ks,
from which we calculate an energy conversion factor of
selected because of the good color information and in
rvl.lxlO-llergs cm-2count-1. order to study the spectral evolution across the achro
3.3. Optical data matic temporal break at ~30 ks. Two X-ray spectra
were produced, the former in the pre-break interval 9-
Figt:re 2 shows the X-ray afterglow light curve, report 20 ks, the latter in the post-break interval 50-1000 ks,
ing the XRT (filled circles) and Chandra (open circles) and scaled to match the observed count-rate at each time
data, and the optical afterglow light curves, including of interest. The two SEDs were jointly fit in count space
data from UVOT, LT, FTS, and SkycamZ. The best fit (Starling et al. 2007) either with a power law or a bro
models are also shown (X-ray: solid line; optical: dashed ken power law continuum. In the latter case the two
lines). spectral slopes were tied so to obey the standard after
The UVOT/ White light curve is well described by glow closure relations. Two dust and gas components,
a broken power law plus a constant that accounts for modeling the Galactic and intrinsic host extinction and
the host galaxy emission. The afterglow initially decays absorption, were also included in the fit. We assumed
with a slope of 0.56±0.04, steepening to 1.57±0.05 af a Solar metallicity for the absorption components and
ter ~29 ks. We estimate a host galaxy contribution of constrained them to the values derived from the XRT
23.4±O.15 mag. spectral fits. We tested three canonical laws - Uilky
A significant afterglow color evolution (~I_B~0.25 Way (MW), Small Magellanic Cloud (SMC), and Large
mag) over the course of the first night was reported l\lagellanic Cloud (LMC) - for the host galaxy extinction
by Haislip et al. (2009). In the fit of the multicolor by using the parameterization of Pei (1992).
light curves we initially allowed for frequency-dependent The resulting fit is shown in Figure 3. Both
slopes and/or temporal breaks, but the sparse sam SEDs are well described (X2=146 for 168 d.o.f.)
pling in the B, V, and Zl filters does not allow us to by a broken power law v.-ith indices .81 =O.300~g:g~o,
detect any color variation. As we found consistent
fh=(31 +0.5=0.800:,:g:g~0 and a decreasing break energy of
results between the different filters, we performed a
joint fit of the BVRr'i'z' light curves by leaving the Ebk=0.15±0.03 keY at 6 ks and Ebk=6:':~ eVat 55 ks. A
normalizations free to vary and tying the other model LMC-type extinction with E(B - V)=0.036±0.015 mag
parameters. The best fit model requires three temporal is only slightly preferred (~X2<2) to a MW-type or a
breaks (X2/d.o.f.=53/70). The model parameters SMC-type law.
6 E. Troja et al.
4. RESULTS
• LongGRB
4.1. Prompt emisSion properties
ShortGRB
4.1.1. Spectml lags GRBSN
GRB noS"
A common property of long GREs is that soft en
ergy photons are delayed with respect to the higher en 1:'
ergy ones. The measurement of such lags is a ,,-alu
~
able tool in the study of G REs and their classification ..J
(e.g. Gehrels et al. 2006). Systematic studies of BATSE 50
and Swift bursts show that long GRBs predominantlr ~
-&
have large, positive lags, ranging from 25 IDS to ",200
s (Norris 2002; Norris et al. 2005; Ukwatta et al. 2010), 48 050509B
while negligible lags are characteristic of short-duration
bursts (Norris & Bonnell 2006; Gehrels et al. 2006) and -,-0v102-11
high-luminosity long GREs (Norris 2002). -2 0 2
The prompt emission of GRE 091127 seems not to log '[3/(1 +z) [sl
fit in this classification scheme. We measured a small
FIG. 4.-Lag-luminosity diagram for long GRBs (squares), short
spectral lag of 731",2.2 ms, consistent with zero, in the GRBs (triangles), nearby GRBs with SNe (filled circ1("s) and with
BAT channels 3-1, and a negative lag of T42~-9.2 ms out SN (open circ1e3). Error bars are 1 0'.
in the BAT channels 4-2. The burst position in the lag
luminosity plane is shown in Figure 4, where we also well above the 95% upper limit derived in § 3.1.2. This
report data for short and long GRBs from the litera is shown in Figure 5, where we report the observed data
ture (Gehrels et al. 2006; McBreen et al. 2008). Having with their best fit Band model extrapolated to the LAT
a negligible lag and only a moderate isotropic peak lumi energy range.
nosity (Lpk,i",~5xlO51 ergss-1), GRE 091127 does not The joint fits reported in Table 1 confirm that
follow the trend of cosmological long GRBs, analogously Fermi/LAT observations are not consistent with the ex
to under-luminous bursts such as GRB 980425. Nearby tension of a Band function from low to high energies, but
sub-energetic bursts (with or without an associated SN) require a steepening of the spectrum at energies below
are outliers of the lag-luminosity relation (thick dashed 100 !\IeV. The inclusion of a high-energy spectral break,
line). The inclusion of GRE 091127 suggests that instead that we modelled as an exponential cut-off, improves the
of simply being outliers, there might be a population fit (Ll.-STAT~38 for one additional degree of freedom).
of bursts following a distinct trend (thin dashed line). Such a break is particularly evident in the K onus-Wind
While a larger sample of nearby bursts is needed to test and in the GBM spectra, and we note that the two fits
this hypothesis, an immediate result coming from Fig yield consistent values of the cut-off energy and an im
ure 4 is that GRB 091127, wh:ch is securely associated provement in the fit statistic of Ll.X2~14 and Ll.X2~9 re
with a massive star progenitor, intercepts the bright end spectively. The quality of the data does not allow us
of the short GRB population, showing that the scatter to constrain the spectral index above the break energy
of long GREs in the lag-luminosity plane is larger than and distinguish between a steepening of the power-law
previously thought. decay or an exponential cut-off. By modeling the high
In the case of GRE 091127, thanks to the GRB low energy data with a simple power-law we derive a photon
redshift and low intrinsic extinction, the associated SN index of ~-3.6, and set an upper limit <-2.6 (90% con
,{as easily revealed by ground-based follow-up observa fidence level). The significance of the high-energy break
tions (Cobb et al. 2010; Berger et al. 2011), nailing down VlaS tested by simulating 10,000 spectra with a simple
the nature of the GRB progenitor. However, had the Band shape. We jointly fit each set of spectra with a
same GRB occured at a higher redshift, its classification Band function (our null model) and a Band function with
would mostly rely on its high-energy properties. At z > 3 an exponential cut-off (the alternative model). The frac
the faint soft emission would be under the BAT detec tional number of simulations in which Ll.-STAT238 gives
tion threshold, and the GRB would appear as a zero lag, the chance probability that a high-energy spectral break
intrinsically short (Tgo/(1 + z) ;:;2 s) burst, similar to improves the fit. None of the simulations showed a varia
GRB 080913 and GRE 090423 for which a merger-type tion of the statistics as high as the one observed, confirm
progenitor was also considered (e.g. Zhang et al. 2009). ing that the presence of a spectral break is statistically
It is also possible that some of the higher redshift short preferred at a >99.99% level.
duration bursts arise from massive star collapses (e.g. The log-parabolic model of IIJassaro et al. (2010) also
Virgili et al. 2011). provides a better fit than the standard Band function
(STAT/d.o.f~679/575 vs. 690/574), and naturally ac
0.2. Softening of the high-en<rgy spectrum
counts for the observed suppression of the high-energy
Fitting results are listed in Table 1 for the time emission.
averaged spectrum and in Tab. 2 for the time-resolved A time-resolved spectral analysis temporally localizes
analysis. By describing the time-integrated spectrum the spectral break during the first 'l'-ray peak (interval
with tile canonical Band function we obtained typical a). In this case the presence of a cut-off at energies "'500-
parameters: a~-1.3, 6~-2.3 and a soft peak energy 1000 keY decreases the flt statistics of ..i.-STAT~25. The
of ~45 keY. However by extrapolating the best fit Band lower significance with respect to the time-averaged anal
model to the LAT energy range, the predicted flux in the ysis is likely due to the lack of Konus-Wind data in
100MeV-IGeV energy band is "'10-7 ergscm-2s-', this fit, however the observed break is evident both in
Broadband study of GRB 091127 7
is LX,i,o~2xlO45ergs-l, very similar to GRB 030329,
and a factor of > 103 brighter than GRB 031203 and other
GRBs/SNe. The UV /optical afterglows appear instead
to decay more rapidly and to cluster at late times but
this could be the result of an observational bias, a~ the
chance of discovering a supernova is higher if the optical
afterglow· is faint.
If the afterglov! emission of GRB 091127 is mainly syn
chrotron radiation from the external forward shock its
broadband behavior has to obey the fireball model 'clo
sure relations (e.g. Zhang & Meszaros 2004). We found
that the GRB afterglow is roughly consistent with a
model of a narrow jet expanding into a homogeneous sur
rounding medium. Our results agree well with previous
studies (Vergani et al. 2011; Filgas et al. 2011). The fire
10 100 1000 10' 10' 10' 10' ball model describes the emission from a population of
Energy [keV] accelerated electrons with energy distribution n( E) ex E-P.
F~G .. 5.- ~es~-fit Band model of the time-averaged spectrum From the afterglow spectral properties we derive an elec
(solid lm~) wIth Its 1 (I confidence interval (dashed lines). Data tron index p=1.60:!:g:~~, which is at the lower end of the
from Swift/BAT, Suzaku-\\'AM, GEM and Kanus-Wind are re p distribution but not uncommon (Panaitescu & Kumar
paroer taeldso w sihtohw tnh.e ir 1 C1 error bars. Upper limits from Fermi/LAT 2002). An achromatic break is detected at ~8hr, after
WhICh the X-ray and optical light curves deca'" with a
similar slope of ""' 1.6. This behavior is suggestive of an
the WAM and in the GBM spectra at a folding energy
early jet-break. The presence of ajel-break at early times
Ecut consIstent between the different instruments. Ac
cording to, this model, the obsen-ed fluence during the IS also supported b~r our Chandra observations, which do
first peak IS (4.3±0.6) X 10-6 erg cm-2 in the 8-1000keV not show evidence of a steepening in the X-ray light curve
several months after the burst. We found that any possi
energy band. At a redshift z~0.49 this corresponds to
~n isotropic equivalent energy E"i,0~(3.5±0.5) x1051 erg ble late time jet-break is constrained to t> 115 d which
for typical parameters, would imply an unusually larg~
In the 1-10,000 keY rest-frame energy band. In this
opening angle OJ >300 Instead the two Chandra points
time interval the derived value of the low-energy index •
hint at a shallower decline, as expected for example in
is a=-0.41:!:g:~8, which is harder but marginally con
the transition to the non-relativistic phase (Piro et a1.
sistent with the limit of 2/3 imposed by the optically
2001).
thin synchrotron emission. The presence of a thermal
The SED analvsis (§ 3.4) shows that optical and X-ray
com parrent is sometimes invoked to explain the hardest
data belong to different branches of the synchrotron
low-energy spectral indices (e.g. Ghirlanda et al. 2003).
As already noted In § 3.1, we tested this hypothesis and sthpeec ttwruom e,n esringcye btahned sc.o oTlihneg ofbresqeruveendc yb rVeca kli east b3e0t wkse eins
found that in no case does the inclusion of a black-bodv
therefore not connected to spectral variations or changes
(single or multi-temperature) yield a significant improv~
in the ambient density. Figure 3 shows that at 6 ks the
ment in the fit statistics, although such a component is
lowest optical flux produced by the X-ray source (with
not inconsistent with the data.
The spectrum of the second peak (interval b) can be I/c just below X-rays) would be only a factor ;S2 lower
than measured, thus the reverse shock contribution to
well descnbed by a Band function. The inclusion of the
the. total optical flux (Kobayashi 2000) is negligible and
LAT data yields a steeper high-energy spectral slope than
optIcal and X-ray emission mainly arise from the same
the one derived from the GBM only fit, and the addition
source (external forward shock). In this framework
of a h:gh-energy break is not required by the data. Ac
the evolution of the cooling frequency is tied to th~
cording to this model, the observed fiuence during this
observed X-ray and optical decays by the following
interval is (4.5±0.2)xlO-6erg cm-2 in the 8-1000keV
relation (Panaitescu et al. 2006):
energy band, corresponding to an isotropic equivalent en
ergy E"iw~(4.3±0.3) x 1051 erg in the 1-10,000 keY rest
d In Vc
frame energy band. - dint ~ 2(Qx-Qopt)=0.94±0.11. (1)
4.2. Afterglow properties
This is co~sistent with our spectral fits, which mea
In Figure 6 we compare the afterglov.' of GRB 091127 to
sure a coolmg frequency that is rapidly moving down
the sample of Swift GRBs with bona fide SN associations
wards in energy as Vc ex t-1.5±O.5, as independently
(Hjorth & Bloom 2011). The observed XRT and UVOT
found in Filgas et al. (2011). For constant microphys
light curves were corrected for red shift and absorption ef
ical parameters, a decreasing Vc sugges:.s an ISM envi
fects, and shifted to a common rest-frame energy band of
ronment rather than a wind-like density profile, where
0.3-10 keY (XRT) and a rest-frame wavelength of 1600 A
the cooling frequency is expected to increase with time.
(UVOT; Oates et al. 2009). From an afterglow perspec
In a uniform density medium, the cooling break evolves
tlve, GRB 091127 resembles the behavior of typical long
-l/2 -3/2t-l/2 h . I
GRBs, dominated by the bright emission from the exter as 1/c ex E EB Clo r a sp enca expansion,
nal forward shock, rather than the unusual evolution of and as Vc ex E-2j3E"B3j2tO in the jet spreading phase
nearby GRBs. The isotropic X-ray luminosity at t~11 hr (Panaitescu & Kumar 2002, 2004). For constant micro-
8 E. Troja et al.
• ,
•
10" DD• 00l033001332120963 D •• 001580001532211599A88 10" \ 060218 ••• 001850001532011299588A
• lJuOfilK • 090618
•
" 980425 • 081007 \ • 08 to07
091127 10" • 091127
10"
-~
::c
:E ...~. 1031
.j;' 10" e.n
!1 ~
.i§
~1030
.3 0
,., Ie" ·"6
e
"
><I .....l 10'"
0«
C)
C)
10" <D
....
...- 10"
10"
100 1000 10' 10' 10' 100 1000 10' 10' 10'
Time since burst [s] Time since burst [sl
FIG. 6.-Left panel: Rest-frame XRT afterglow lightcurvcs for Swift GRBs with an associa.ted SN (filled circlE'S). We also report the da.ta
for the three pre-Swift bursts with a spectroscopicaJl~' confirmed SN (open symbols, Kaneko et aI. 2007). Right Panel: Reost-frame UVOT
afterglow light curves. Only GRBs with an UVOT detection arc shown. Early time LT /FTS data for GRB 091127 are also reported.
physical parameters and no energy injection into the Opt:eal depth effects - The lack of high-energy photons
blastwave, the expected decay is shallower than the ob in bright bursts such as GRB 091127 could be an indica
served one. This shows that the simplest version of the tion of a pair opacity break (Beniamini et al. 2011), and
fireball model can not account for the overall afterglow therefore used to constrain the outflow Lorentz factor
behavior I and, as we vIill discuss in Sect. 5.2, some mod· (Lith~'ick & Sari 2(01). In order to be self-consistent
ifications (e.g. energy injection or evolving microphysca.l these calculations rely on the fundamental assumption
parameters) are required. that the observed sub-MeV spectrum extrapolates to
aeV energies. Following this line of argument, we can
5. DISCUSSION
set a first upper limit on the bulk Lorentz factor in
5.1. Origin of the high-energy speetml break aRB 091127 just by considering its non-detection by
The most recent Fermi observations of aRBs sug LAT. We use here the Band function parameters and
impose E ",<100MeV:
gested that the prompt '1'-ray emission can be satisfacto m
rily described by a smoothly broken power law, the Band
function, extending to the GeV energies, often accom [C~~7:,V) t;-l].m
r" < 130 !loo (2)
panied b.v an additional non-thermal component mod
eled as a power law (Zhang et al. 2(11). In this burst
we found that a standard Band function, though pro where ,6=2.28 is the high-energy spectral slope and
viding an adequate description of the spectrum in the flOo ~O.l ph cm-2 S-1 keV-1 the observed flux den
keY energy range, is in contrast with the simultaneous sity at 100 keY, both derived from the spectral fit in
Fermi/LAT observations as it overpredicts the observed Tab. 1. The variability timescale was set to tv"'0.3 s,
emission above 100MeV (see Fig. 5). The spectral fits the minimum value observed in the "I-ray lightcurve.
presented in Table 1 and Table 2 detect the presence of a In deriving Eq. 2 we approximated the spectral shape
spectnl softening at ",0.5-1 MeV in the time-integrated with a simple power-law, fv <X v-p. Given that for
spectrum and during the first peak of emission. This E=100MeV the typical energv of the target photons is
disfavors spectral evolution as the origin of the observed E, ~ 1 MeV> > Ep'" the effects of low-energy spectral
feature. curvature (Baring & Harding 1998) can be considered
A steepening of the high-energy spectral slope could be negligible and a simple power-law decay is a valid ap
caused by several factors such as absorption from the ex proximation.
tragalactic background light (EBL), attenuation via pair The upper limit derived in Eq. 2 is based on the simple
production (')''1' --+ e±) or an intrinsic break in the en formulation given in Lithwick & Sari (2001), where spa
ergy distribution of the emitting electrons. Based on the tial and temporal dependencies are averaged out. More
low redshift of this burst (z=0.49) and the low energy realistic calculations taking into account the progressive
of the observed break, EBL absorption can be excluded buildup of the radiation field further decrease the above
(see e.g. Finke et al. 2010). Belov! we consider in turn value by a factor of 2-3 (Hascoet et a1. 2011), that is
the other possibilities. r n '" 50. This is significantly lower than the values
Broadband study of GRB 091127 9
estimated for cosmological GRBs (Molinari et al. 2007; 'J"GeV /'J"keY ;S 0.01, disfavors Synchrotron Self Comp
Liang et al. 2010), though similar to the Lorentz factor ton as the main radiation mechanism (Piran et a1. 2009;
inferred for X-ray flares (Abdo et al. 2011). If we now Beniamini et al. 2011).
take into account the observed steepening at ;S1 MeV The observed spectrum of the first ,-ray peak, is
as it c:-iginates from an increase in the optical depth, by roughly in agreement with a fast cooling synchrotron
setting Emax. ~Ecut we get r ~ 2. Such a low Lorentz spectrum: the low-energy slope a~-O.4 is marginally
factor, tr"ough atypical for classical GRBs, is not un consistent with the maximum slope of 2/3 allowed for
precedented (Soderberg et al. 2006). A weakly relativis v<v" while the high-energy slope fJ ~-2 suggests that
tic outflow could therefore account for the lack of high for E >50 keV we are already above the injection fre
energ~j photons, and the observed soft spectrum, but not quency Vm. It follows that VcR::Vm, that is the ef
for the bright afterglow detected a few minutes after the fects of adiabatic and radiative cooling are comparable
burst. (marginal fast cooling; Daigne et al. 2011). In the ex
An independent estimate of the bulk Lorentz factor treme case r c/r m I"V 10, synthetic spectra resemble the
can be derived from afterglow observations. The dura observed spectral shape: a hard low-energy tail followed
tion of the GRB being rather short, we consider the thin by a smooth, flat transition (vm<v<lO vm) to the fi
shell case (Kobayashi et al. 1999). Since the afterglow nal Fv'X v-p/2 decay. The observed steepening at
is already fading in our first observation we can assume "'0.7 MeV from fJ~-2 to <-2.6 corresponds to this
that be onset happened at tpk <140 (l+z)-1 ~100rso, transition, and implies p ~ 3.2. However when the slow
and set a lower limit to the outflow Lorentz factor cooling contribution is significant, the radiative efficiency
(Piran 1999): decreases markedly (Daigne et al. 2011), and it is hard
) to account for the high luminosity and variability of the
ro > 240 ( E,,~; 1/8 (3) prormcpirt emission. If we consider the more efficient case
1]0.2 n pk,2 of m ~1, then the spectral break has to be ascribed
to a different mechanism.
where E;,=1052 E-y,52 erg is the isotropic-equivalent . A spectral cut-off is expected at vbM), where '"(M'
energy, TJ = O.21}!J.2 is the radiative efficiency and is the maximum Lorentz factor of the shocked elec
n ~1 cm-3 is the medium density (Bloom et al. 2003). trons. Such a break occurs at energies ;:;200 l\IeV
By using the empirical relation suggested by Liang et al. (Bosnjak et al. 2009), and it is unlikely at the origin of
ro
(2010), we infer a similar high value of ~200. the MeV break. An alternath,-e explanation is an intrin
The limits derived from the prompt and afterglow sic curvature in the energy distribution of the radiating
emission properties are inconsistent: the former suggest electrons (Hassaro et al. 2010), arising if the higher en
a mildly relativistic outflow (r < 50, or even r "'2), ergy electrons are accelerated less efficiently than those
the latter a highly relativistic jet (r » 100). A possi with lower energy.
bilitv that would reconcile the two sets of limits is that
the fiBt spectrally softer pulse, during which we detect 5.2. Jet collimation and energetics
the significant presence of a spectral break, is instead
From our broadband spectral fits of the prompt
the GRB precursor originating at RR::2r2ct R:: lOll em,
u emission we derived an isotropic equivalent energy
e.g. from the jet cocoon emerging from the progenitor
E"i'o =(1.1±0.2) x 10'2 erg, which is in the typical range
star (Lazzati & Begelman 2005). A different physical ori
of long GRBs (Bloom et al. 2003). The afterglow prop
gin could also explain the different lags between the two
erties show evidence of a tightly collimated outflow, indi
main '"V-ray events and the unusual lag evolution: while
cating that the true energy release is significantly lower.
spectral lags in GRB pulses generally tend to increase
The achromatic nature of the break at tbk ~30 ks and the
with time (Hakkila et al. 2008), it has been found that
subsequent afterglow fast decay are typical signatures of
precursors have larger lags than the following 'Y-ray emis
a jet-break, and we first consider this hypothesis. ill this
sion (Page et al. 2007). Precursors, however, carry only a
scenario the jet opening angle OJ is:
small fraction of the total energy release (Morsony et al.
2007), while the first peak encloses 50% of the observed
e. =4.2 (Ei'0"2)-1/8 (tbk )3/8 deg, (4)
')'-ray fiuence.
We therefore are led to consider that our assumption J 'Ia.2n 8 hr
of a palr opacity break is not valid, that is: 1) the and the collimation-corrected energy is
incomistency between Eq. 2 and Eq. 3 implies that the E"j=(3.0±0.8) x 1049 erg, at least an order of mag
Band-t.ype spectrum does not extend to Ge V energies, nitude lower than typical long GRBs (Cenko et al.
but a spectral break (not related to optical depth effects) 2010). However, as noted in § 4.2, this simple fireball
below 100 MeV. is required by the data; 2) we identify scenario fails to reproduce two main features: 1) the
this b,eak with the steepening at ",0.7 MeV, which is rapid temporal evolution of the cooling frequency; 2) the
therefore an intrinsic feature of the GRB spectrum. observed pre-break flux decay rates (ax = 1.03±0.04,
aopt = 0.56±0.04), which are not compatible with
Breaks in GRB spectra - We discuss here the standard the model expectations (av>v" 1"V0.7, av,,>v>vm 1"V0.45
scenario, in which internal shocks within the expanding for a spreading jetj av>vc = av,,>v>vm I"VO.8 for a
outflow accelerate the ambient electrons to relativistic non-spreading jet; Panaitescu & Kumar 2004). In order
energies with a power-law distribution n(E) ex: E-P. The to reconcile the observed afterglow behavior with the
GRB prompt emission originates as synchrotron radi theoretical expectations one needs to invoke either a
ation from the shock-accelerated electrons. The small continual energy injection and/or evolving microphysi
ratio between the GeV and keV fluences of this burst, cal parameters. The former scenario would require an
10 E. Troja et al.
limated outflow (OJ;S4'), a low prompt 'Y-ray energy
(E, <3x 1049 erg), and a total relativistic energy yield
;0: : ~o of Em1 ;S3 X 1050 erg, at the lower end of the long GRBs
distribution. In Figure 7 VIe compare the burst energetics
;:: ~ with the sample of long GRBs. Independently from the
~ 031203 Long GRBs afterglow model adopted (I, narrow jet + energy injec
10 - L tion: star; II, evolving €B: diamond), the' burst location
~ ~6•021 8
in the lower left corner shows that GRB 091127 more
~ .030329
w - ~n. ~XRF closely resembles the class of X-ray Flashes (XRFs) and
c 091127 020903 GRBs/SNe rather than typical GRBs. This is also con
- sistent with its rather soft spectrum and unusual lags.
'c" 98~/ 091127 (I)
6. SUMMARY AND CONCLUSION
We presented a broadband analysis of the prompt and
0.1 10 lOa afterglow emission of GRB 091127, securely associated
EK [10'" erg] with SN2009nz. Two main features emerged from our
FIG. 7.-:- Prompt emission" energy release, E'l' versus afterglow study of the prompt emission: I) the burst is character
kinetic "lnergy, EK. For GRB 091127 both the scenarios discussed ized by small, negligible spectral lags; 2) the high energy
in the text are shown: (I) narrow jet + prolonged energy injec (>IOOMeV) emission is significantly suppressed. The
t(ifoilnle; d (cIiIr)c leevso),l vXinRgF s~ B(.o peWn ec irrecpleosr),t adnadt ab uforrs tss twaintdha ar dsp leocntrgo sGcoRpBics GRB has a long duration (T90~7s), and a relatively soft
SN (squares). Values are corrected for collimation effects. Ref: spectrum (EpkR<45 ke V). However having negligible spec
Panaitucu & Kumar (2002); Bloom et al. (2003); Soderberg et aI. tral lags and only a moderate luminosity, the burst does
(2004); Cenko et aI. (20lO). not fit the lag-luminosity relation followed by cosmolog
ical long GREs, but lies in the region of short duration
extreme injection episode, the jet energy increasing by a
bursts. While the association with SN2009nz leaves no
factor of 100 in the first 8 hours. Furthermore, there is
douMs about the origin of the GRB progenitor, the atyp
no apparent reason for the injection to end at the time
ical lag behavior adds additional uncertainty in the clas
of the jet-break, leading to an even larger shock energY
sification of GRBs based solely on their high-energy prop
carried by the slower ejecta. The alternative possibility
erties. It also links GRB 091127 to nearby sub-energetic
of a growing magnetic energy fraction € B is discussed by bursts, such as GRB 980425, which are also outliers of
Filgas et al. (2011).
the lag-luminosity relation.
~-e found instead that a narrow confined jet, 'Vihose By modeling the GRB prompt emission with the
boundary is visible from the first afterglow measurement
(i.e. r < 8;1), and a prolonged energy injection, last standard Band function, we found that such a model
significantly overpredicts the observed flux at higher
ing until rv30 ks, provide a consistent description of the
(>100 MeV) energies. Consistently, our spectral fits
afterglow temporal and spectral properties and ease the
show eyidence of a spectral curvature at energies
energetic burden without requiring any variation of the
;S11\[eV. If due to opacity effects, the suppression of high
shock microphysical parameters. For an ISM-like circum
energy emission would suggest a low outflow Lorentz fac
burst medium (§ 4.2), the flux decay indices are given by tor (r < 50, or even r ~ 2), as measured in nearby sub
(Panaitescu & Kumar 2004):
energetic GREs. However, this interpretation is not con
3 p+4 sistent with our early-time detection of a bright fading
"'0 = 4P - -4-e, (5) afterglow, which suggests r»100. We therefore con
clude that the high-energy break is an intrinsic property
3p+ I p+3
"'x = -4- - -4-e, (6) of the GRB spectrum.
The multi-wavelength afterglow emission is character
where e is the power-law evolution of the forward-shock ized by an achromatic break at ",8 hr after the burst, and
energy E ex te. The above set of equations overconstrain by a rapidly decaying cooling frequency, lie ex: t-1.5±0.5.
the e parameter, thus providing a consistency check of We considered two scenarios to interpret these features
the solution. By substituting in Eq. 5 and 6 the ob within the standard fireball model. The former interprets
served pre-break temporal slopes and the value of p~1.6 the achromatic break as ajet-break, from which we derive
from the broadband spectral fit, we derive e=0.48±0.06 a jet opening angle 8j ~4 deg, and a collimation-corrected
and e=0.39±0.06, respectively. Departures of the energo" energy E,R<3xl049 erg. This. model needs to let the mi
injection from a pure power law can explain the optical crophysical parameters vary with time in order to re
plateau at t<5 ks, while the cessation 6f energy injection produce the observed temporal decays and the rapidlv
at R<3C ks yields the observed achromatic break. Accord decreasing lie. The latter scenario instead interprets the
ing to this model, by imposing tbk < 140s in our Eq. 4 achromatic break as the end of a prolonged energy injec
we derive OJ ;S0.6 (nil cm-3)l/S deg, and E, ;S6x1047 tion episode, the jet-break happening before the start of
(nil cm-3)l/4erg. By using e ~0.45, the blastwave ki our observations (t<140 s). According to this model, we
netic energy can be constrained to &<. ;S3x 1050 erg, most derive a jet opening angle OJ ;S 0.6 deg, and a collimation
of which comes from the slower ejecta that are gradually corrected energy E,;S6xI047 erg. This GRB therefore
replenishing the forward shock energy. presents hybrid properties: a high luminosity ')'-ray emis
From our analysis the following features clearly. sion powered by narrowly collimated and highly rela
emerge: GRB 091127 is characterized by a highly col- tivistic outflow as typical of long GREsj its low-energy