An achromatic break in the afterglow of the short GRB 140903A: evidence for a narrow jet
E. Troja, T. Sakamoto, S. B. Cenko, A. Lien, N. Gehrels, A. J. Castro-Tirado, R. Ricci, J. Capone, V. Toy, A. Kutyrev, N. Kawai, A. Cucchiara, A. Fruchter, J. Gorosabel, S. Jeong, A. Levan, D. Perley, R. Sanchez-Ramirez, N. Tanvir, S. Veilleux
aa r X i v : . [ a s t r o - ph . H E ] J un . Draft version June 17, 2016
Preprint typeset using L A TEX style emulateapj v. 5/2/11
AN ACHROMATIC BREAK IN THE AFTERGLOW OF THE SHORT GRB 140903A:EVIDENCE FOR A NARROW JET
E. Troja , T. Sakamoto , S. B. Cenko , A. Lien , N. Gehrels , A. J. Castro-Tirado , R. Ricci , J. Capone ,V. Toy , A. Kutyrev , N. Kawai , A. Cucchiara , A. Fruchter , J. Gorosabel † , S. Jeong , A. Levan ,D. Perley , R. Sanchez-Ramirez , N. Tanvir , S. Veilleux Draft version June 17, 2016
ABSTRACTWe report the results of our observing campaign on GRB 140903A, a nearby ( z = 0 . T ∼ Swift . We monitored the X-ray afterglow with
Chandra up to 21 days after the burst, and detected a steeper decay of the X-ray flux after t j ≈ θ j ≈ E ≈ × erg. We further discuss the nature of the GRB progenitorsystem. Three main lines disfavor a massive star progenitor: the properties of the prompt gamma-rayemission, the age and low star-formation rate of the host galaxy, and the lack of a bright supernova.We conclude that this event was likely originated by a compact binary merger. Subject headings:
X-rays: bursts; gamma ray burst: individual (GRB 140903A); INTRODUCTION
Gamma-ray bursts (GRBs) are produced by a highlyrelativistic outflow collimated into jets. The angular sizeof the outflow is therefore a key ingredient in determin-ing the true energy release and the event rate. These Department of Astronomy, University of Maryland, CollegePark, MD 20742, USA NASA, Goddard Space Flight Center, 8800 Greenbelt Rd,Greenbelt, Greenbelt, MD 20771, USA Department of Physics and Mathematics, College of Scienceand Engineering, Aoyama Gakuin University, 5-10-1 Fuchinobe,Chuo-ku, Sagamihara-shi, Kanagawa 252-5258, Japan Joint Space-Science Institute, University of Maryland,College Park, MD 20742 Department of Physics, University of Maryland, BaltimoreCounty, Baltimore, MD 21250, USA Instituto de Astrof´ısica de Andaluc´ıa (IAA-CSIC), P.O. Box03004, E-18008 Granada, Spain Unidad Asociada Departamento de Ingenier´ıa y SistemasAutom´aticos, E.T.S. Ingenier´ıa Industrial, Universidad deM´alaga, Campus de Teatinos, Arquitecto Francisco Penalosa, 6,29010, M´alaga, Spain INAF-Istituto di Radioastronomia, Via Gobetti 101, I-40129Bologna, Italy 0000-0003-4631-1528 Department of Physics, Tokyo Institute of Technology,2-12-1 (H-29) Ookayama, Meguro-ku, Tokyo 152-8551, Japan Space Telescope Science Institute, 3700 San Martin Drive,Baltimore, MD 21218 Unidad Asociada Grupo Ciencias Planetarias (UPV/EHU,IAA-CSIC), Departamento de F´ısica Aplicada I, E.T.S. Inge-nier´ıa, Universidad del Pas Vasco (UPV/EHU), Alameda deUrquijo s/n, E-48013 Bilbao, Spain. Ikerbasque, Basque Foundation for Science, Alameda deUrquijo 36-5, E-48008 Bilbao, Spain , Universidad del Pa´ısVasco, Bilbao, Spain. Sunkgkyunkwan University, 25-2 Sungkyunkwan-ro,Jongno-gu, 1398 Seoul, Korea. Department of Physics, University of Warwick, Coven-try,CV4 7AL, UK Dark Cosmology Centre, Niels Bohr Institute, University ofCopenhagen Juliane Maries Vej 30, 2100 Copenhagen, Denmark Department of Physics and Astronomy, University ofLeicester, Leicester, LE1 7RH, UK † Deceased parameters provide a crucial test for any progenitor andcentral engine model.Measuring the collimation of short duration GRBs,i. e. those lasting less than 2 s (Kouveliotou et al.1993), is not only a primary interest of the GRB field,but has a broader impact. Growing observational evi-dence connects short GRBs with compact binary merg-ers (Gehrels et al. 2005; Tanvir et al. 2013; Berger 2014;Yang et al. 2015, and references therein), which areamong the most promising sources of gravitational wave(GW) radiation (Thorne 1987; Abbott et al. 2016a).Therefore, the degree of collimation of short GRBs is acritical input for inferring the true rate of binary mergers,the expected detection rate of advanced LIGO and Virgo(Abadie et al. 2010), and for estimating our chances toobserve the electromagnetic counterpart of a GW source(Abbott et al. 2016b; Troja et al. 2016).Observationally, the beamed geometry leaves a clearsignature in the afterglow temporal evolution, manifest-ing itself as an achromatic light curve break (known as“jet-break”), visible on timescales of ∼ days-weeks afterthe explosion (Rhoads 1999). At early times (hours afterthe explosion), the evolution of the afterglow is the sameas for a spherical explosion. However, later on, the jetedges become visible causing the observed flux to rapidlyfall off (van Eerten et al. 2010; van Eerten & MacFadyen2013). For a jet expanding into a homogeneous ambientmedium such steepening takes place at a time t j ∝ θ / j (Sari et al. 1999; van Eerten et al. 2010), when the out-flow is decelerated down to a bulk Lorentz factor ≈ θ − j ,where θ j is the jet half-opening angle. The detection ofa jet-break in the afterglow light curve is therefore animportant diagnostic tool for constraining the outflowgeometry, and the burst energetics.In the case of short bursts, the faintness of their af-terglows often hampers the search for jet-breaks. Onlya small fraction of short GRBs have been detected at E. Troja et al.optical or radio wavelengths, and often sampled toopoorly to meaningfully constrain the afterglow tempo-ral evolution (Kann et al. 2011; D’Avanzo et al. 2014).Nicuesa Guelbenzu et al. (2012) presented good evidencefor an achromatic steepening in the optical/NIR lightcurve of the short GRB 090426. However, the classifi-cation of this burst is rather ambiguous (Antonelli et al.2009; Levesque et al. 2010), and it was proposed thatthe event was more likely an interloper, originated by amassive star progenitor (Th¨one et al. 2011; Virgili et al.2011; Nicuesa Guelbenzu et al. 2012)Candidate jet-breaks have been identified in severalX-ray afterglows of short GRBs (Burrows et al. 2006,Soderberg et al. 2006, Stratta et al. 2007; Fong et al.2012; Coward et al. 2012; Zhang et al. 2015), howevertheir interpretation as jet-breaks remain quite contro-versial. Several studies suggest that the X-ray lightcurves may be shaped by a persistent energy injectionfrom the central engine (Fan & Xu 2006; Cannizzo et al.2011; Rowlinson et al. 2013) rather than by externalshock emission (M´esz´aros & Rees 1997). In this scenariothe sharp decay of the X-ray flux is attributed to therapid turn-off of the energy source rather than to theoutflow geometry, and no collimation is needed to ex-plain the observed light curves. Indeed, in the smallsample of events with simultaneous optical and/or ra-dio coverage (e.g. GRB090510, De Pasquale et al. 2010;GRB130603B, Tanvir et al. 2013) the observed temporalbreaks appear to be chromatic rather than frequency-independent. The jet-break interpretation, to still hold,would require an alteration of the basic jet model, such asa two-component jet (Corsi et al. 2010), evolving shockparameters (De Pasquale et al. 2010), or the presence ofadditional emission components (e.g. Gao et al. 2015).In this paper, we present our multi-wavelength cam-paign of the short GRB 140903A which revealed anachromatic break in its afterglow light curve. Throughthe analysis of the broadband data, we show that theobserved emission is fully consistent with the standardforward shock model, and requires a narrowly collimatedoutflow. A previous analysis of this event, based on Swift observations, did not detect the presence of a jet-breakin the X-ray data (Fong et al. 2015). Our addition ofdeep, late-time
Chandra observations is indeed criticalfor the jet-break detection and its characterization. Wefurther investigate the GRB classification and the natureof its progenitor, and conclude that this event is a bonafide short GRB, likely originated by a compact binarymerger. The paper is organized as follows: our observa-tions and data reduction procedures are detailed in § §
3; our results are discussedin §
5. Throughout the paper, times are referred to the
Swift trigger time, and the phenomenology of the burst ispresented in the observer’s frame. We employ a standardΛCDM cosmology with H = 67 . − Mpc − , Ω M =0.308, and Ω Λ =0.692 (Planck Collaboration et al. 2015).Unless otherwise stated, errors are given at the 68% con-fidence level for one interesting parameter, and upperlimits are reported at the 3 σ confidence level. OBSERVATIONS AND DATA REDUCTION
Swift BAT and XRT
GRB 140903A triggered the
Swift
Burst Alert Tele-scope (BAT; Barthelmy et al. 2005) at 15:00:30 UT on3rd September, 2014 (Cummings et al. 2014). The
Swift
X-ray Telescope (XRT; Burrows et al. 2005) began set-tled observations of the GRB field 74 s after the BATtrigger, and monitored the X-ray afterglow during thefollowing 3 days, until the source faded below the detec-tor sensitivity threshold. The XRT data comprise 24 ksacquired in Photon Counting (PC) mode.BAT and XRT data were processed using the
Swift software package distributed within HEASOFT (v. 6.17).We used the latest release of the BAT and XRT Cali-bration Database, and followed standard data reductionprocedures.
Chandra
The
Chandra
X-Ray Observatory performed two Tar-get of Opportunity (ToO) observations in order 1) to pre-cisely localize the X-ray afterglow (PI: T. Sakamoto), and2) to characterize its late-time temporal evolution, andsearch for a possible jet-break (PI: E. Troja). Our firstobservation (ObsId 15873) started 3 d after the burst,and observed the field for a total exposure of 19.8 ks.Our second
Chandra observation (ObsId 15986) was per-formed on 2014, Sep 18. for a total exposure of 59.3 ks.
Chandra data were reduced using version 4.6.1 of theCIAO software with CALDB version 4.6.3. Events fromthe GRB afterglow were selected using a source extrac-tion radius of 2 pixels, and the derived count rates werecorrected for vignetting effects and Point Spread Func-tion (PSF) losses. The background contribution was es-timated from an annular, source-free region centered onthe afterglow position.The GRB afterglow is detected at both epochs. Inour first
Chandra observation we detect 80 net sourcecounts in the 0.5-8.0 keV energy band. We correctedthe native
Chandra astrometry by aligning our X-rayand optical images (see § α = 15 h m . s , δ = +27 ◦ ′ . ′′
83 with an error radius of 0.4 ′′ (90% con-fidence level). In our second and last Chandra obser-vation only 6 counts are measured at the source posi-tion, corresponding to a detection significance > Discovery Channel Telescope
We initiated an observing campaign with the LargeMonolithic Imager (LMI) mounted on the 4.3 m Discov-ery Channel Telescope (DCT) in Happy Jack, AZ. Ob-servations in the griz filters started on 2014 Sep 04 at3.17 UT, approximately 12 hours after the
Swift trig-ger, and continued to monitor the field for the next 3weeks. Late-time images in the r and i filters were ac-quired on 2016 March 17 (561 days after the burst) andused as templates for image subtraction. Standard CCDreduction techniques (e.g., bias subtraction, flat-fielding,etc.) were applied using a custom IRAF pipeline. In- IRAF is distributed by the National Optical Astronomy Ob-servatory, which is operated by the Association of Universities for hort GRB jet-break 3
NE 10 arcsecChandraXRT promptXRTenhanced
Fig. 1.—
DCT r -band observations of the field of GRB 140903A, taken at 0.5 days ( left panel ) and 2.5 days ( middle panel ) after theburst. The black circle shows the initial XRT afterglow localization. The blue and red circles show the XRT enhanced and the refined Chandra positions, respectively.
Right panel : image subtraction of the two previous panels, showing the residual afterglow light. dividual short (10–20 s) exposures were aligned with re-spect to astrometry from the Sloan Digital Sky Survey(SDSS; Ahn et al. 2014) using SCAMP (Bertin 2006) andstacked with SWarp (Bertin et al. 2002).As shown in Figure 1, the field of GRB 140903A isquite complex: the optical afterglow lies on top of a rela-tively bright host galaxy (see below), and only 12 ′′ awayfrom an extremely bright ( V ≈ t = 12 . . σ ).Using the images from 2016 March 17, we measurethe following magnitudes for the underlying host galaxy: r ′ = 20 . ± .
09, and i ′ = 20 . ± .
05. From ear-lier observations we also measure g ′ = 21 . ± .
16, and z ′ = 19 . ± .
08, although we caution that these fluxesmay include some afterglow contribution. The host is un-resolved in all of our DCT images (seeing ranging from0 . ′′ . ′′ α = 15 h m . s , δ = +27 ◦ ′ . ′′
68. The excess af-terglow flux measured in our subtracted images is consis-tent with this location, within the estimated uncertaintyof our astrometric tie ( ≈
100 mas in each coordinate).
Liverpool and Calar Alto Telescopes
Near-IR images were acquired in zJHK s -bands usingthe 2.0m Liverpool (LT) and the 3.5m Calar Alto tele-scopes (CAHA). The LT images were taken in the z -bandwith the IO:O camera, which provides a 10 . ′ × . ′ fieldof view and a 0 . ′′ pixel scale. The CAHA data wereacquired in the JHK s -bands with the Ω instrument,yielding a 15 . ′ × . ′ field of view and a 0 . ′′ pixel scale.In order to reduce the contamination of the nearby brightstar, these observations were taken in relatively short (20s - 30 s) exposures. The reduction followed standardsteps; bad pixel masking, bias and flat field correction,sky subtraction, plus stacking, performed by calling onIRAF tasks (Tody 1993). The resulting photometry, cal- Research in Astronomy (AURA) under cooperative agreement withthe National Science Foundation. ibrated with respect to nearby point sources from SDSSand 2MASS (Skrutskie et al. 2006), is presented in Ta-ble 1. We used the offsets from Blanton & Roweis (2007)to convert the 2MASS Vega magnitudes to the AB sys-tem.
Gemini Imaging
We imaged the field of GRB 140903A with the Gem-ini Multi-Object Spectrograph (GMOS; Hook et al. 2004on the 8 m Gemini North telescope. We obtained a sin-gle 120 s i ′ image beginning at 05:24 UT on 2014 Sep 4(∆ t = 14 . ×
60 s i ′ exposures at a mean epoch of ∆ t = 39 . gemini IRAF package. We performed digital image subtractionon the GMOS images using the same analysis methodsas was used for the DCT images (Sect. 2.3). In the sub-tracted frame transient emission is clearly detected atan offset of 96 ±
44 mas from the galaxy’s center. Ata redshift z =0.351 this corresponds to a physical pro-jected offset of 0.5 ± i ′ = 21 . ± .
05 mag in our first epoch,and i ′ = 22 . ± .
13 mag in the second epoch. Thisimplies a steep temporal decay with slope α o =1.54 ± Gemini Spectroscopy
We obtained a series of spectra of the afterglow+galaxywith GMOS beginning at 05:34 UT on 2014 Sep 4 (∆ t =14 . Fig. 2.—
Gemini GMOS spectrum of GRB 140903A and its hostgalaxy, acquired 14.6 hrs after the burst. The positions of detectedemission and absorption lines are indicated. Crossed circles markthe position of strong telluric features.
E. Troja et al.
TABLE 1Log of Optical and Near-IR Observations
Date Time Since Burst Telescope Instrument Filter Exposure Time Afterglow Magnitude a Host Magnitude a (UT) (d) (s) (AB) (AB)2014 Sep 4.13 0.51 DCT LMI r ′
300 21 . ± . · · · r ′
300 21 . ± . · · · i ′
120 21 . ± . · · · r ′ > . · · · i ′
600 22 . ± . · · · r ′ > b · · · i ′ > . · · · g ′ · · · ± z ′ · · · ± z ′ > . · · · r ′ > · · · i ′ > . · · · z ′ · · · ± J · · · ± H · · · ± K s · · · ± i ′ > . · · · r ′ · · · ± i ′ · · · ± i ′ · · · ± a Values not corrected for Galactic extinction. b A faint excess ( r =23.11 ± . σ ), and we cannot excludethat it is an artifact of the subtraction method. and a central wavelength of 600 nm, providing cover-age from λ ≈ ≈ λ> ≈
15 ˚A) absorption line at λ ≈ ≈
10 ˚A)absorption line at λ ≈ I with z ≈ .
35. We alsodetect narrow emission lines at λ = 6569 . ± . λ = 6763 . ± . β and [O III ]at z = 0 . ± . . Weak absorption features corresponding toCa II H+K are also visible at this redshift, though withmarginal significance.
GTC Spectroscopy
Further optical spectroscopy of the host galaxy wasperformed using OSIRIS (Optical System for Imagingand low Resolution Integrated Spectroscopy; Cepa et al.2000) at the 10.4 m GTC. Observations started on Oct03, 2014, i.e. ∼ ×
600 s exposures) and R2500I VPH (3 ×
600 s ex-posures). The spectra covered the 3600–7800 ˚A range ata resolution of ≈ ≈ × The weaker [O
III ] λ z =0.351. TABLE 2Log of Radio Observations
Date Time Since Burst Frequency Flux(UT) (d) ( GHz) ( µ Jy)2014 Sep 04.06 0.44 6.1 118 ± ± ± ± ± < < were flux calibrated using the spectrophotometic stan-dard star GD248, which was observed during the samenight with a 2.52 ′′ slit. In order to account for slit losses,we renormalized the flux of the source to match the DCTmagnitudes shown in Table 1. Acquisition images werenot usable due to the nearby saturated star.Although close to a skyline, H α is clearly detected inthe red spectrum at λ = 8862 . ± . Jansky Very Large Array
GRB 140903A was observed with the Jansky VeryLarge Array (VLA) at both 6.1 GHz (C-band) and at9.8 GHz (X-band). Observations started ∼
10 hrs af-ter the burst, and periodically monitored the sourcefor 18 days (Fong et al. 2015). Radio data were down-loaded from the public NRAO archive, and reduced usingthe Common Astronomy Software Applications (CASA)v. 4.5.2 package. After standard calibration and ba-hort GRB jet-break 5sic flagging, we visually inspected the data and appliedfurther screening when needed. Galaxies 3C286 andJ1609+2641 were used as flux and phase calibrators,respectively. The log of radio observations is reportedin Table 2. Our values are slightly higher, but largelyconsistent with those reported by Fong et al. (2015).A simple power-law fit to the data yields decay slopes α GHz =0.63 +0 . − . and α . GHz > t> DATA ANALYSIS
Gamma-ray data
The prompt emission consists of a main Fast RiseExponential Decay (FRED) pulse, with a duration of T = 0.30 ± χ =44 for 57 degrees of freedom) by a simplepower-law with Γ=1.99 ± +0 . − . ) × − erg cm − , which, at aredshift z =0.351, corresponds to an isotropic-equivalentenergy of E γ, iso =(6.0 ± × erg. Due to the nar-row BAT energy bandpass, this only places a lower limitto the bolometric energy release. However, for a typi-cal GRB spectrum (Band et al. 1993), the measured softphoton index indicates that the spectral peak lies close toor within the BAT energy range (Sakamoto et al. 2009).In this case, the bulk of the emission mainly falls withinthe observed range, and the derived value of E γ, iso rep-resents a good estimate of the total energy radiated inthe prompt emission.Spectral lags were calculated by cross-correlating thelight curves in the standard BAT channels: 1 (15-25keV), 2 (25-50 keV), 3 (50-100 keV), 4 (100-350 keV). Wefollowed the method outlined by Ukwatta et al. (2012)and, in order to increase the signal-to-noise in the higherenergy channels, performed the analysis on non mask-weighted lightcurves, each with a 4 ms time resolu-tion. The derived lags are τ =12 +7 − ms and τ =-1 +7 − ms,where the quoted uncertainties were derived by MonteCarlo simulations. The results of our lag analysis are Fig. 3.—
Left panel : BAT light curve of GRB 140903A in the 15–150 keV energy band. The T time interval, and the time intervalused for the cross-correlation function (CCF) analysis are shown. Right panel : CCFs between the standard BAT energy bands. Thebest fit gaussian function is reported as a solid line. The lag valueand its uncertainties are indicated by the vertical shadowed region. shown in Fig. 3 (right panel).We also searched for temporally extended emis-sion following the main burst, but no significant sig-nal was found. By assuming a power-law spectrumwith photon index Γ=2, we set a 3 σ upper limit of8 × − erg cm − s − (15-50 keV) in the time interval10-100 s. This is consistent with the MAXI upper limitof 8.4 × − erg cm − s − in the 4-10 keV energy band(Serino et al. 2014). X-ray data
Spectral analysis
The afterglow spectral parameters were derived fromthe time-averaged XRT/PC spectrum (from 100 s to110 ks). We binned the data in order to have at least1 count per spectral channel, and performed the fitwithin XSPEC (Arnaud 1996) v.12.9.0 by minimizingthe Cash statistic. The spectrum is well described byan absorbed power law model (W-stat=329 for 359 de-grees of freedom, d.o.f.). The best fit parameters are aphoton index Γ X =1.66 ± N H , int ( z =0.351)=(1.3 ± × cm − , in excess to theGalactic value N H , Gal =2.9 × cm − in the burst di-rection (Kalberla et al. 2005). The adopted value is con-sistent within errors with the N H , Gal =3 . × cm − estimated by Willingale et al. (2013).In our first Chandra observation, the source spec-trum is well fit (W-stat=55 for 58 dof) by anabsorbed power law model with Γ X =1.8 ± N H , int =1.3 × cm − , fixed at the value of the XRTbest fit. In our second and last Chandra observation thelow number of counts prevents any spectral analysis. Asthe hardness ratios of the two
Chandra observations areconsistent within the uncertainties, the same spectral pa-rameters were adopted to estimate the observed flux.For the best fit parameters quoted above, we de-rived an unabsorbed energy conversion factor (ECF) of(4.8 ± × − erg cm − cts − for the Swift /XRT data,and of (1.40 ± × − erg cm − cts − for the Chan-dra
ACIS-S data.
Temporal Analysis
The X-ray light curve was binned to have a minimumof 15 counts in each temporal bin. The observed count-rates were converted into flux units by using the ECFsderived in Section 3.2.1, and by propagating the relativeuncertainties. We modeled the afterglow temporal de-cay with a series of power-law segments ( f X ∝ t − α i ) andminimized the χ statistics to obtain the best fit to thedata. The afterglow displays a shallow decay phase withtemporal index α ∼ α ∼ t bk, ∼ Chandra /ACIS-S data point liesbelow the predictions based on the
Swift /XRT dataset,hinting at a second temporal break in the light curve.However, the combined XRT/ACIS-S dataset could bereasonably well described by adopting a steeper tem-poral index α ∼ Chandra observation wastherefore executed in order to distinguish between thetwo models. This last measurement confirms the pres-ence of an additional break in the X-ray light curve at atime t j ≈ α ∼ Fig. 4.—
Afterglow light curves of GRB 140903A, combining X-ray data from the
Swift /XRT (small circles), and the
Chandra /ACIS-S(large circles), optical data from DCT (open squares), Gemini (filled squares), and radio data from the VLA (diamonds). Error bars are 1 σ ,arrows denote 3 σ upper limits. The best fit temporal model is shown as a solid line. The vertical band marks the time of the jet-break. summarized in Table 3. The X-ray light curve and ourbest fit model are presented in Fig. 4, and compared tothe optical (Table 1) and radio measurements (Table 2)in order to highlight the achromatic nature of the lasttemporal break t j , which we interpret as the jet-breaktime. TABLE 3Afterglow light curve fit parameters
Band α t bk, α t bk, α χ / dof(ks) (ks)X 0.20 ± +0 . − . +0 . − . +17 − +0 . − .
43 / 46O 1.54 ± ± +1 . − . +0 . − . +11 − +0 . − .
49 / 48R (6.1 GHz) -0.5 a +11 − ± > a The temporal slope was held fixed at the value predicted by thestandard fireball model for ν sa <ν<ν m . Afterglow Spectral Energy Distribution
In order to study the spectral evolution across the tem-poral break t j detected in X-rays, we extracted the after-glow spectral energy distribution (SED) at two differentepochs, t =0.5 d ( < t j ) and t =2.5 d ( > t j ). These timeswere selected in order to maximize the simultaneous cov-erage at different wavelengths. Optical fluxes were derived by the best fit temporalmodel in Table 3, and corrected for Galactic extinc-tion in the GRB direction ( E B − V ≈ f ν ∝ ν − β ) to the optical andX-ray data yields spectral slopes β OX =0.72 ± t = t , β OX =0.76 ± t = t , significant intrinsic ab-sorption N H =(1.8 ± × cm − , and marginal ev-idence of dust extinction A V = 0 . ± .
25. The sim-ple power-law fit provides a good description of thedataset (W-stat=355 for 371 d.o.f.), suggesting that op-tical and X-ray emission belong to the same spectralsegment ( ν m < ν opt < ν X < ν c ) of the synchrotron spec-trum. The lack of significant spectral variation acrossthe temporal break t j is consistent with the properties ofa jet-break, and exclude alternative interpretations (e.g.cooling frequency).By extrapolating the observed spectrum to radio en-ergies, the predicted flux at t = t is ≈
10 mJy, two or-ders of magnitude higher than the radio measurement.This implies a spectral break between the optical andradio band, and that the radio data belong to a differentspectral segment ( ν r <ν m ). By adopting the standardclosure relations for GRB afterglows (Zhang & M´esz´aros2004) we fixed the radio spectral index to β r = 1 / Fig. 5.—
Left panel:
Afterglow spectral energy distribution at two different epochs, t =0.5 d before the jet-break, and t =2.5 d after thejet-break. We fit the broadband spectrum with a smoothly broken power-law, our best fit models are shown by the solid lines. The thindotted lines show the effects of absorption and extinction. Top right panel:
Temporal evolution of the peak frequency across the jet-break.We report the expected behavior for three different models: the spherical fireball (dotted line), a narrow jet ( θ jet =0.1 rad) seen on-axis(dashed line), and seen slightly off-axis ( θ obs /θ jet =0.6; solid line). Our measurements, indicated by the red diamonds, agree well with theoff-axis jet model. Bottom right panel:
Same as above but for the spectral peak flux. Also in this case, our derived values (cyan circles)agree well with the trend expected from an off-axis jet. of interstellar scattering and scintillation at radio wave-lengths. Although a proper estimate of the ISS fluctua-tions requires more complex modeling (see Section 4.2),at this stage we introduce an uncertainty of ≈ ν pk ≈ × Hz at 0.5 d. At our second epoch, theradio measurements are only slightly lower than the ex-trapolation of the higher energy spectrum, implying thatthe spectral peak moved close to the radio band. Weestimate ν pk ≈
37 GHz at 2.5 d, above the VLA frequen-cies. This shows that the observed radio, optical andX-ray emission remained in the same spectral regime,thus the observed temporal break was not caused byspectral variations. Basic considerations on the spec-tral and temporal behavior of the afterglow disfavor awind-like environment, which would cause a steeper de-cay ( α wind ≈ ν pk ∝ t − , and f pk ∝ t − . . As shown in Fig-ure 5 (right panels), these decay rates are significantlysteeper than the ones predicted by the spherical fireballmodel for a uniform medium, and are instead consistentwith the spectral evolution of a collimated outflow. Inparticular, the slow decay of the peak flux strongly favorsa narrow jet model seen slightly off-axis. Host Galaxy Properties
GRB 140903A is located on top of a compact andred galaxy, suggestive of an old system. Based on thegalaxy sky densities in the r -band (Yasuda et al. 2001),we estimated a small probability of a chance association, P ch ≈ × − (Bloom et al. 2002; Troja et al. 2008), andwe therefore consider this galaxy as the GRB host. Fromour r -band measurement we derive a rest-frame absolute B -band magnitude M B ≈ -20.9 mag, or L B ≈ . L ∗ whencompared to the luminosity function of galaxies at a sim-ilar redshift 0.2
Photometry of the galaxy hosting GRB 140903A. Data(filled circles) are corrected for Galactic extinction in the directionof the GRB. The best fit stellar population synthesis model (graycurve), and its parameters are reported. lar mass log(M/M ⊙ )=10.61 ± t =4.1 +3 . − . Gyr, and a moderate star formation rateSFR =1.0 ± ⊙ yr − in agreement with the presenceof nebular emission lines in our spectra.By using the extinction corrected H α line flux we in-fer a comparable value of SFR = 0.38 ± ⊙ yr − (Kennicutt 1998) for a Chabrier IMF. and a specific SFRof 0.47 ± L/L ∗ ) M ⊙ yr − . Based on the diagnos-tic F ([O III ] λ F ( Hβ ) ∼ ≈ ± RESULTS
Origin of the X-ray emission
The early X-ray afterglow of GRB 140903A is char-acterized by a period of fairly constant emission last-ing ≈ α ∼ directly powered by the cen-tral engine. One of the most popular models invokes anewborn magnetar as the power source of the GRB andits afterglow: as the magnetar spins down, it injects en-ergy into the jet causing a period of nearly flat emission(the plateau), followed by a steeper temporal decline withslope α & ν a , we can derive a rough esti-mate of the emitting radius R & × cm at t =0.5 d(Barniol Duran et al. 2013), consistent with an externalshock origin. Moreover, the observed temporal and spec-tral indices ( β OX ≈ α ≈ ν m < ν X < ν c and p ≈ indirectly poweredby the central engine via sustained energy injection intothe forward shock and, after the cessation of energy in-jection is communicated to the shock front, the afterglowevolves in a standard fashion (van Eerten 2014). There-fore, the X-ray emission is not directly linked to the timehistory of the central engine, instead it carries impor-tant information about the jet collimation, energetics,and surrounding environment. Afterglow modeling
We modeled the broadband dataset (from radio toX-rays) by using the standard prescriptions for an ex-panding spherical fireball, and the scaling relations forthe post-jet-break evolution (Sari et al. 1999). We ex-cluded from the fit the early time data ( t < t bk, )as they are affected by persistent energy injection. Inour fit we implemented a routine to calculate the ex-pected ISS modulation for each set of input after-glow parameters. By adopting the ‘NE2001’ model(Cordes & Lazio 2002), we derived a scattering measureSM = 1.3 × − kpc/m − / and a transition frequency ν =8 GHz in the direction of GRB 140903A. Observa-tions below this frequency could possibly be affected bystrong scattering if the source size is smaller than the ISSangular scale, θ F ≈ µ as. At the GRB redshift this corre-sponds to an apparent fireball size R ⊥ . × cm, whichis likely the case at the early timescales here considered.The derived ISS fluctuations were treated as a source ofsystematic uncertainty and added in quadrature to thestatistical errors when evaluating the fit statistics.We assumed a uniform circumburst medium with den-sity n , and constant microphysical parameters ǫ e and ǫ B . Under these assumptions, we did not find an accept-able fit to the data ( χ =65 for 43 dof), mainly becausethe model predicts a much faster decay of the peak fluxand peak frequency after the jet-break. We attemptedto model this effect by leaving the microphysical parame-ters free to vary in time as ǫ e ∝ t e and ǫ B ∝ t b . Althoughhort GRB jet-break 9the fit formally improves for b ≈ e ≈ ǫ e >
1, and extreme values forthe blastwave kinetic energy and the jet opening angle.We considered this model an unrealistic description ofthe explosion, and turned to a different interpretation toexplain the observed properties.As shown in Figure 5 (right panels), the temporalevolution of the broadband spectrum appears roughlyconsistent with a collimated fireball observed slightlyoff-axis. We therefore introduced in our model theeffects of different viewing angles (van Eerten et al.2010; van Eerten & MacFadyen 2013). This providesa better description of the observed data. Thebest fit parameters are an isotropic equivalent ki-netic energy E K , iso =4.3 +1 . − . × erg, a circumburstdensity n =0.032 +0 . − . cm − , and shock parameters ǫ B =2.1 +3 . − . × − , ǫ e =0.14 +0 . − . . We derived a jet open-ing angle of θ j =0.090 ± θ obs ≈ Constraints on SN-like transients
The possibility of an optical/IR transient rising afew days after the short GRB explosion is the cur-rent focus of intense research (e.g. Barnes & Kasen 2013;Yu et al. 2013; Kasen et al. 2015). The detection andidentification of such transients (e.g. Tanvir et al. 2013;Yang et al. 2015; Jin et al. 2015, 2016) would representthe smoking gun proof of short GRB progenitors, and apowerful tool to search for electromagnetic counterpartsof GW sources. We used our late-time observations toconstrain some of the most promising models as well asthe presence of an emerging supernova.As shown in Figure 7 (left panel) our r -band upperlimits at 2.5 d and 4.5 d can constrain the presenceof a fast-rising and rapidly decaying transient, peak-ing in the optical a few days after the burst. Weconsidered two models: the classical Li & Paczy´nski(1998) macronova (or kilonova) powered by the ra-dioactive decay of the ejecta (shaded area), and themore recent merger-nova (Yu et al. 2013) powered by along-lived magnetar (shaded area). Recent theoretical(Barnes & Kasen 2013) and observational (Tanvir et al.2013) results showed that the macronova emission isheavily suppressed at optical wavelengths due to thehigh-opacity of the ejecta. Models for the late-time in-frared emission (e.g. Barnes & Kasen 2013), althoughhighly dependent on the input physics, generally predicta signal ( H &
23 mag at t ∼ M ej =0 . ⊙ (thin solidline) we can exclude only the extreme values of the f parameter ( f > × − ), which measures the fraction ofradioactive material converted into heat. Our limit is more interesting in the case of a larger ejecta mass of M ej =0.1M ⊙ , for which we can exclude f > − (thicksolid line). This is consistent with the most recent calcu-lations of radioactive heating rate (Metzger et al. 2010;Lippuner & Roberts 2015).Yu et al. (2013) argued that, if the GRB central en-gine is a stable magnetar, the macronova luminositycould be boosted by several orders of magnitude. Inthis scenario, the main power source is the magnetar-driven wind rather than the radioactive decay energy. Asshown in Figure 7, for a typical range of ejecta masses( M ej . − M ⊙ ) the predicted signal of a merger-nova(dashed line) could be consistent with our observations.As mentioned in Section 2.3, we found marginal ( . σ )evidence of a signal in our observations 2.5 days postburst. The resulting magnitude, r =23.11 ± Gemini ob-servations at 1.5 d and the DCT observations at 2.5 d.When compared with the macronova predictions, thissignal would require either an extreme value of the f -parameter 5 × − We have presented several lines of evidence linkingGRB 140903A to the class of short duration GRBs(Kouveliotou et al. 1993), and in support of the popu-lar compact binary merger model. Although character-ized by a rather soft spectrum with photon index Γ ∼ T ∼ L γ, iso ∼ erg s − ), all key features of the classof short GRBs (Norris & Bonnell 2006; Gehrels et al.2006). The GRB afterglow was found on top of a rela-tively bright galaxy. Given the accurate afterglow local-ization, the probability of a chance alignment can be con-sidered negligible ( P ch ≈ Fig. 7.— Left Panel: Late-time r -band upper limits compared with theoretical light curves of a macronova (solid lines) and a magnetar-driven merger-nova (shaded area). The dash-dotted symbol shows the low-significance signal visible in our DCT image at 2.5 days. Themacronova signal was derived by using the following parameters: a lanthanide-free opacity κ =1 cm g − , ejecta velocity v =0.1 c , ejectamass M ej =0.01 M ⊙ and a rather high radioactive energy deposition f =2 × − (thin solid line); κ =1 cm g − , v =0.1 c , M ej =0.1 M ⊙ ,and f =10 − (thick solid line). The merger-nova model was calculated by assuming a long-lived stable magnetar, and ejecta masses10 − M ⊙ Late-time i -band observations compared with the extinction-corrected template light curves of GRB-SNe: SN1998bw (solid line), andSN2006aj (dashed line). j ≈ θ j ≈ ◦ , an isotropic-equivalentenergy release E K , iso ≈ × erg, and a viewing angle θ obs ≈ ◦ . Our modeling yields a blast-wave kinetic en-ergy that is significantly higher than the observed promptgamma-ray energy. This would imply an unusually lowradiative efficiency, η γ ≈ f b ∼ 250 has a direct impact on theGRB energy release and true event rate, and therefore onthe progenitor models. Coward et al. (2012) estimate theobserved rate of short GRBs as ∼ − yr − . Collima-tion can boost this number up to ∼ × Gpc − yr − ,which is consistent with the conservative rate density ofNS-NS mergers from Abadie et al. (2010). This wouldsuggest that most NS mergers successfully launch a shortGRB, and that other systems, such as NS-BH or whitedwarf binaries, do not contribute significantly to the ob-served GRB population. An important caveat to theabove comparison between observations and progenitormodels is that estimates of GRB jet angles are unavoid-ably biased by our observing strategy and limited sensi-tivity. Narrowly collimated jets, if pointed toward us, aremore likely to trigger Swift over a larger volume and toproduce bright afterglows, allowing for the jet-break de-tection. On the other hand, wide outflows of comparableenergy produce dimmer GRBs and afterglows, which areharder to detect and characterize. A proper assessmentof the GRB event rate should properly account for theseobservational biases.The collimation-corrected energy release ishort GRB jet-break 11 E ≈ × erg, which is in the typical range forshort GRBs and lower than average long durationbursts (Cenko et al. 2010; Zhang et al. 2015). Recently,Perna et al. 2016 proposed a new mechanism to powera short GRB from a BH-BH collision. However, the lowdisc mass available in this system could only power afaint, low-luminosity transient, not consistent with theenergetics measured in our case. GRB 140903A wasmore likely produced by a merger event in which atleast one of the two compact objects was a neutron star.According to the standard NS merger model, a stellar-mass black hole surrounded by a hot massive torusis formed after the merger. Energy is extracted fromthis system through neutrino anti-neutrino annihilationor magnetically driven mechanisms. Pair annihilationof neutrinos and antineutrinos can supply an energydeposition rate L ν ¯ ν . erg s − (Setiawan et al. 2004;Birkl et al. 2007), consistent with the energy budget ofGRB 140903A. Following the formalism of Fan & Wei(2011), we use the burst energetics to estimate a post-merger disc mass M disc ≈ ⊙ . This is in agreementwith numerical simulation of merging NS-NS and NS-BHbinaries. If instead the outflow is driven by more efficientmagnetic processes, the disc mass could be as low as10 − M ⊙ , suggesting a high-mass binary NS merger(Giacomazzo et al. 2013). An alternative scenario is theformation of a supra-massive and highly magnetized neu-tron star after the merger (Giacomazzo & Perna 2013).In this case, there are less robust predictions connectingthe central engine and the GRB observed properties.A general requirement is that the total energy releaseshould not exceed the maximum rotational energy ofthe newborn NS, E rot ≈ ( M NS /2 M ⊙ ) / erg. Theburst energetics are well below this limit, and consistentwith the proto-magnetar model. A compact binarymerger can therefore naturally explain the observedGRB properties, although the nature of the centralengine and the energy extraction mechanisms remainuncertain. Only future detections of gravitational waveradiation will be able to ultimately discriminate betweenthese different scenarios. CONCLUSIONS We detected a temporal break in the X-ray afterglowlight curve of the short GRB 140903A. The afterglowtemporal decay was observed to steepen from α ∼ α ∼ t j ≈ × erg. Severallines of evidences link this event to the popular NSmerger scenario: the prompt gamma-ray emission, theenvironment, the lack of a bright SN, the energetics andrate of events. Our results show that NS mergers canproduce highly collimated outflows.The scientific results reported in this article are basedin part on observations made by the Chandra X-ray Ob-servatory. Support for this work was provided by theNational Aeronautics and Space Administration throughChandra Awards GO4-15072A and GO4-15067A issuedby the Chandra X-ray Observatory Center, which is oper-ated by the Smithsonian Astrophysical Observatory forand on behalf of the National Aeronautics Space Ad-ministration under contract NAS8-03060. These resultsalso made use of Lowell Observatory’s Discovery Chan-nel Telescope. Lowell operates the DCT in partnershipwith Boston University, Northern Arizona University, theUniversity of Maryland, and the University of Toledo.Partial support of the DCT was provided by Discov-ery Communications. LMI was built by Lowell Obser-vatory using funds from the National Science Founda-tion (AST-1005313). This paper is partly based on ob-servations obtained at the Gemini Observatory, whichis operated by the Association of Universities for Re-search in Astronomy, Inc., under a cooperative agree-ment with the NSF on behalf of the Gemini partnership:the National Science Foundation (United States), the Na-tional Research Council (Canada), CONICYT (Chile),Ministerio de Ciencia, Tecnolog´ıa e Innovaci´on Produc-tiva (Argentina), and Minist´erio da Ciˆencia, Tecnologiae Inova¸c˜ao (Brazil). Observations were also carried outwith the 10.4 m Gran Telescopio Canarias installed in theSpanish Observatorio del Roque de los Muchachos of theInstituto de Astrofisica de Canarias in the island of LaPalma (GTC59-14B) and with the 3.5m CAHA telescopeat the German-Spanish Calar Alto Observatory operatedby the IAA-CSIC. AJCT acknowledges support from theSpanish Ministry Projects AYA2012-39727-C03-01 and2015-71718R. REFERENCESAbadie, J., Abbott, B. P., Abbott, R., et al. 2010, Classical andQuantum Gravity, 27, 173001Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2016a, Phys.Rev. Lett., 116, 061102Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2016b, ArXive-printsAhn, C. P., Alexandroff, R., Allende Prieto, C., et al. 2014, ApJS,211, 17Aloy, M. A., Janka, H.-T., & M¨uller, E. 2005, A&A, 436, 273Antonelli, L. A., D’Avanzo, P., Perna, R., et al. 2009, A&A, 507,L45Arnaud, K. A. 1996, in Astronomical Society of the PacificConference Series, Vol. 101, Astronomical Data AnalysisSoftware and Systems V, ed. G. H. Jacoby & J. Barnes, 17Band, D., Matteson, J., Ford, L., et al. 1993, ApJ, 413, 281 Barnes, J. & Kasen, D. 2013, ApJ, 775, 18Barniol Duran, R., Nakar, E., & Piran, T. 2013, ApJ, 772, 78Barthelmy, S. D., Barbier, L. M., Cummings, J. R., et al. 2005,Space Sci. Rev., 120, 143Bauswein, A., Goriely, S., & Janka, H.-T. 2013, ApJ, 773, 78Becker, A. 2015, HOTPANTS: High Order Transform of PSFANd Template Subtraction, Astrophysics Source Code LibraryBerger, E. 2014, ARA&A, 52, 43Bertin, E. 2006, in Astronomical Society of the Pacific ConferenceSeries, Vol. 351, Astronomical Data Analysis Software andSystems XV, ed. C. Gabriel, C. Arviset, D. Ponz, & S. Enrique,1122 E. Troja et al.