Mon.Not.R.Astron.Soc.000,1–??(2011) Printed9January2012 (MNLATEXstylefilev2.2) The origin of the early time optical emission of Swift ⋆ GRB 080310 2 O.M. Littlejohns1,† R. Willingale1, P.T. O’Brien1, A.P. Beardmore1, S. Covino2, 1 0 D.A. Perley3, N.R. Tanvir1, E. Rol4, F. Yuan5, C. Akerlof6, P. D’Avanzo2, 2 D.F. Bersier7, A.J. Castro-Tirado8, P. Christian3, B.E. Cobb9, P.A. Evans1, n a A.V. Filippenko3, H. Flewelling10, D. Fugazza2, E.A. Hoversten11, A.P. Kamble4 J 5 S. Kobayashi7, W. Li3, A.N. Morgan3, C.G. Mundell7, K. Page1, E. Palazzi12, ] R.M. Quimby13, S. Schulze14, I.A. Steele7, A. de Ugarte Postigo15 E H 1 Department of Physics and Astronomy, Universityof Leicester, LE1 7RH, UK 2 INAF/Osservatorio Astronomico di Brera, via Emilio Bianchi 46, 23807 Merate (LC), Italy . h 3 Department of Astronomy, Universityof California, Berkeley, CA 94720-3411, USA p 4 Astronomical Institute “Anton Pannekoek”, P.O. Box 94248, NL-1090 SJ Amsterdam, The Netherlands - 5 Research School of Astronomy and Astrophysics, The Australian National University, CotterRoad, Weston Creek, ACT 2611, Australia o 6 PhysicsDepartment, Universityof Michigan, AnnArbor, MI48109, USA r 7 Astrophysics Research Institute, Liverpool John Moores University,Twelve Quays House, Egerton Wharf, Birkenhead, CH41 1LD, UK t s 8 Institutode Astrofisica de Andaluca (IAA-CSIC), P.O. Box 03004, E-18008, Granada, Spain a 9 Department of Physics, The George Washington University,Corcoran 105, 725 21st St, NW, Washington, DC 20052, USA [ 10 Institute for Astronomy, University of Hawaii at Manoa, Honolulu, HI96822, USA 1 11 Department of Astronomy & Astrophysics, The Pennsylvania State University,525 Davey Laboratory, UniversityPark, PA 16802, USA v 12 INAF - IASF di Bologna, viaGobetti 101, 40129 Bologna, Italy 2 13 Cahill Centerfor Astrophysics 249-17, California Institute of Technology, Pasadena, CA 91125, USA 9 14 Centre for Astrophysics and Cosmology, Science Institute, University of Iceland, Dunhagi 5, 107 Reykjavk, Iceland 2 15 Dark Cosmology Centre, NielsBohr Institute, Universityof Copenhagen, Juliane MariesVej 30, 2100 Copenhagen, Denmark 1 . 1 0 2011Dec31 2 1 : v ABSTRACT i We present broadband multi-wavelength observations of GRB 080310at redshift z = X 2.43. This burst was bright and long-lived, and unusual in having extensive optical r and near IR follow-up during the prompt phase. Using these data we attempt to a simultaneouslymodeltheγ-ray,X-ray,opticalandIRemissionusingaseriesofprompt pulses and an afterglow component. Initial attempts to extrapolate the high energy modeldirectly to lowerenergiesforeachpulse revealthat a spectralbreakis required between the optical regime and 0.3 keV to avoid over predicting the optical flux. We demonstrate that afterglow emission alone is insufficient to describe all morphology seen in the optical and IR data. Allowing the prompt component to dominate the early-timeopticalandIRandpermittingeachpulsetohaveanindependentlowenergy spectral indices we produce an alternative scenario which better describes the optical light curve. This, however, does not describe the spectral shape of GRB 080310 at early times. The fit statistics for the prompt and afterglow dominated models are nearly identical making it difficult to favour either. However one enduring result is that both models require a low energy spectral index consistent with self absorption for at least some of the pulses identified in the high energy emission model. Key words: gamma-rays:bursts. ⋆ Based on observations made also with ESO Telescopes at the La Silla and Paranal Observatory under programme IDs 080.D- 0250and080.D-0791 2 O.M. Littlejohns et al. 1 INTRODUCTION GRB 080310, which begin at a time soon after the trigger thatisnotoftenaccessedbyground-basedfacilities.Follow- Over the last few years a combination of fast-response ing the trigger, GRB 080310 was detected on-board Swift ground-based telescopes triggered by the availability of by both the X-Ray Telescope (XRT) (Burrows et al. 2005) rapid,accuratelocalisationshavestartedtoprovidethedata and UV/Optical Telescope (UVOT) (Rominget al. 2005) requiredtoanswerthequestionofwhatiscausingtheearly, and also observed in the optical and IR by several ground- bright X-ray and optical emission from gamma-ray bursts based telescopes. These rare data present us with an op- (GRBs). The most accurate prompt X-ray locations come portunity to discriminate between whether the early time fromtheSwiftsatellite(Gehrels et al.2004).Thesearesup- lower-energy light curve of a GRB is driven by internal or plemented by either on-board or ground detections of the external emission, during which time the high-energy emis- ultraviolet (UV),optical or infra-red (IR)counterpart. sion is presumed to be totally internally dominated. In thepopularrelativistic fireballmodel for GRBs, the In the following section we discuss the observations, early, usually highly variable emission is understood to be thenin§3wepresentattemptstofitbothaninternalshock duetointernalshocks(Sari & Piran1997)ormagneticdissi- model(Genet & Granot 2009)andan afterglow component pationwithinthejet,andtheso-calledexternalemission is (Willingale et al. 2007) to the X-ray and γ-ray data, and producedbytheinteraction ofthejet withthesurrounding describe the necessary modification required to simultane- medium. The latter emission is usually described using the ously fit the early optical emission in two scenarios, where fireball model (Rees & Meszaros 1992), which has success- either the prompt or afterglow components dominate this fully been applied to describe the behaviour of GRBs half early flux.In §4 we briefly discuss therelative merits of us- a day or so after the trigger, but has difficulties explaining ing each component, before finally in §5 we conclude which the complex behaviour seen in the first few hours, a period of the two alternatives is a better fit to the observed data nowroutinelyaccessedbySwiftandotherrapid-responsefa- and consider the implications of the model on the physics cilities. Ideallymulti-wavelengthobservationsshouldbeob- governing theemission from GRB 080310. tainedwhiletheburstishappeningsoastotrytodisentan- gle the relative contribution from the internal and external components. Evans et al. (2009) present a uniformly analysed com- prehensive sample of 317 Swift GRBs spanning from De- 2 OBSERVATIONS cember 2004 to July 2008, in which their morphologies On 2008 March 10 the Swift BAT triggered and lo- are compared to the proposed canonical X-ray light curve cated GRB 080310 (trigger number 305288) on board at (Nousek et al. 2006, Zhang et al. 2006 & Panaitescu et al. 08:37:58 UT (Cummings et al. 2008). Swift slewed immedi- 2006). Such canonical light curves consider the X-ray emis- ately which enabled the narrow field instruments to begin sion to consist of a series of power-laws, where one impor- observing the burst 89s after the trigger. The burst was tant phase is the rapid decay phase which has been ex- detected by the XRT and UVOT (white filter), with the plained as being the smooth continuation of the prompt latter providing the best Swift position of RA(J2000) = emission (O’Brien et al. 2006, Tagliaferri et al. 2005 & 14h40m13s.80, Dec(J2000) = −00◦10′29′.′6 with a 1σ error Barthelmy et al. 2005). From the sample of Evans et al. ′′ radius of 0.6. Figure 1 shows a UVOT v-band image from (2009) it is clear that the X-ray light curves of GRBs vary theearlytimedata.TheSwiftlightcurvesobtainedinmul- from bursttoburst.Someshowstrongflaring,oneexample tiple bands from each of the three on board instruments beingGRB061121(Page et al.2007),however,somebursts arepresentedinFigure2andmostoftheavailabledatasets show remarkably simple and smooth decay (GRB 061007; from an extensive number of facilities are shown in Figure Schadyet al. 2007). Similar findings are also reported in 3.Datafrom thePAIRITEL(Perley et al.2008)andKAIT Racusin et al. (2009). Rapid behaviour, such as flaring, at instrumentsare shown separately in Figure 4. highenergiesisoftenattributedtocentral-enginebehaviour InadditiontoSwiftobservations,andthoseinstruments (Margutti et al. 2010) but how this relates to the optical already mentioned, GRB 080310 was also observed on the emission remains somewhat of a mystery. The available ground with numerous optical and NIR facilities, including datasets reveal a confusing picture. In some cases the early REM (Covino et al. 2008), VLT (Covino et al. 2008) and opticaldataseem totracetheX-rayandγ-raylightcurves, theFaulkesTelescope North.These observations are shown such as GRB 041219A (Vestrand et al. 2005 & Blake et al. alongside theBAT and XRTlight curvesin Figure 3. 2005),suggestingthatopticalflaringmaybeofinternalori- The Kast dual spectrometer at the Lick Observatory, gin. In other GRBs the optical behaviour seems entirely California, obtained the first redshift estimation for this unrelated to the the high-energy emission (GRB 990123; burst of z = 2.4266 (Prochaska et al. 2008) using strong Akerlof et al.1999&GRB060607A;Nysewander et al.2009 absorption features from Silicon, Carbon and Aluminium. &Molinari et al.2007),andinstead seemstofollow thebe- This was later corroborated by the VLT-UVES instrument haviourof theexternal afterglow. (Vreeswijk et al.2008)andtheKeck-DEIMOSspectrometer To make progress requires continued efforts to observe (Prochaska et al. 2008). GRBs over as wide a wavelength range as possible from The following subsections describe the observations in as early as possible. This is only really viable for bright, moredetail.TheSwiftdataanalysiswasperformedusingre- long-lasting GRBs which are well-placed for rapid follow- lease2.8oftheSwiftsoftwaretools.Parameteruncertainties up.Herewepresentprompt,multi-wavelengthdatafromthe areestimatedatthe90%confidencelevel.Wenotethatthe opticaldatasetshavebeenreducedusingdifferentmethods, † E-mail:[email protected](OML) and fully investigated the effects of cross calibration errors, The early time emission of GRB 080310 3 Figure1.UVOTv-bandimagefromT0+79stoT0+1539s.The enhancedSwiftXRTpositionisshown(Osborneetal.2008),with anerrorcircleofradius1′.′4,correspondingtothe90%confidence limit, containing the optical afterglow candidate at an average v-bandmagnitudeof17.49. ensuringthatourlateranalysisremainedinsensitivetotheir effects. Figure 2. Swift prompt light curves from GRB 080310 in dif- ferent energy bands. From top to bottom, BAT 100−150keV, 2.1 BAT BAT 50−100keV, BAT 25−50keV, BAT 15−25keV, XRT 1.5−10.0keV,XRT0.3−1.5keVandUVOTvandwhite bands. The BAT data were processed using the standard bat- Strictly,thecountratesforBATareperfullyilluminateddetec- grbproductscript.ThetopfourpanelsofFigure2showthe tor. BATlight curvesdisplayedinthestandardenergybandsof 15−25keV, 25−50keV, 50−100keV and 100−150keV, plottedwithrespecttotheBATtriggertime(T0).Thebin- T0+95stoT0+799sandphotoncountingmodedataafter ningissuchthateachbinsatisfiesaminimumsignal-to-noise this (Beardmore et al. 2008). As can be seen in Figure 2 ratio of 5 and a minimum time bin size of 1s. thesoft band XRTlight curveinitial fades until T0+130s, Theγ-raylightcurveshowsmanypeakswiththefirstat before theburst rebrightens. This flaring activity continues T0−60s.ThebrightestpeakextendsfromT0−12stoT0+7s. untilT0+420sbefore thecount rate dropsagain briefly.A Thisisfollowedbyaperiodofnodetectableemissionbefore furtherflaringeventisseenbetweenT0+500sandT0+620s, a weaker, broad series of peaks is seen from T +180s to which is approximately half as bright as the first. There is 0 T +360s(Tueller et al.2008).Thelatterpeakisconsistent significantspectralevolutionduringtheflaringevents,where 0 with thefirststrong flareseen in theXRT(seebelow). The theS(1.5−10keV)/S(0.3−1.5keV)ratioshowsahardening BATemission isstrongest inthelowerenergybands,below inthespectrumatT0+135stoT0+200sandT0+500sto ∼100keV.TheT90 isestimated tobe365±20s(wherethe T0+565s before, in both instances, there is a softening as error includes systematics). theflaring behaviour declines. ThetotalspectrumfromT −71.76stoT +318.75sis 0 0 well fit by a power-law of photon index 2.32±0.16, with a totalfluenceof2.3±0.2×10−6ergcm−2overthe15−150keV 2.3 UVOT band. The fluence ratio S(25−50keV)/S(50−100keV) is UVOT observations began with a 100 second finding chart 1.27±0.17whichputsGRB080310 ontheborderoftheX- exposuretakenatT +99s(Hoversten & Cummings2008). ray-richgamma-rayburstsandX-rayflash(XRF)according 0 Thisfindingchartexposurewasthesumof10individual10 to thedefinition of Sakamoto et al. (2008). second exposures, with GRB 080310 being detected in all but the first exposure. At the time of GRB 080310 UVOT observations during the first orbit were taken in the event 2.2 XRT mode, which allows for higher timing resolution, while dur- The Swift XRT started observations of GRB 080310 89s ing subsequent orbits observations were taken in imaging after the trigger, with windowed timing data ranging from mode.UVOTphotometrywasdoneusingthepubliclyavail- 4 O.M. Littlejohns et al. 10-7 101 10-8 100 XRT BAT 10-9 FTN R x 0.1 10-1 1) REM R x 0.1 s(cid:0) P60 R x 0.1 2 10-10 ROTSE R x 0.1 10-2 m(cid:0) Super-LOTIS R x 0.1 x (erg c10-11 ILGNTeT-mR WAinTFiCC-Na r+m' Gx r M0' .xO1 S0 .r1' x 0.1 10-3 cal (Jy) V flu10-12 FPT6N0 iI' xx 00..33 10-4 Opti e LT-RATCam i' x 0.3 10k10-13 VLT V 10-5 3- UVOT v x 0.02 0.10-14 UUVVOOTT wb hx x0 .00.0052 10-6 UVOT u x 0.002 UVOT w1 x 0.005 10-15 REM H x 0.5 10-7 WHT H x 0.5 10-16 VLT H x 0.5 10-8 100 101 102 103 104 105 106 time since BAT trigger (s) Figure 3.γ-raytonearIRlightcurvesforGRB080310,showingtheenergybandsusedtoobservetheGRBandtheinstrumentsused totakethemeasurements.Datahavebeenscaledfromtheiractual valuesforviewingpurposes. able FTOOLS data reduction suite, and is in the UVOT Johnson-Cousins system.1 Photometry was then performed photometricsystemdescribedinPoole et al.(2008).During using an aperture matched to the average seeing of the the first 1,000 seconds the light curve is complex as shown (combined) frames. For the conversion from magnitude to inFigure2,afterwhichthetheburstcanbeseentofadein flux,thedatawerefirstcorrectedforGalacticextinctionus- all theobserved optical bands(Figure 3). ing the COBE-DIRBE extinction maps from Schlegel et al. There are available data for the UVOT white, v, b, u (1998), and then converted using flux zero points from anduvw1.Asthisbursthasameasuredredshiftofz = 2.43 Fukugitaet al. (1995) for optical and Tokunaga & Vacca it is important to consider absorption from the intervening (2005)forinfrared.ABmagnitudeswereconvertedfollowing medium through which the emission must travel. Correc- Oke& Gunn (1983). tionfactorswerecalculatedforLymanabsorptionusingthe GRBz code described in Curran et al. (2008). GRBz uses the model presented by Madau (1995) to calculate the ab- 2.5 ROTSE sorptionfromneutralHydrogenintheintergalacticmedium. ROTSE-IIIb, located at McDonald Observatory, Texas, re- Having found the correction factor for all the optical and spondedtoGRB080310 andbegan imaging 5.7seconds af- nearinfraredbandspresentedinthiswork,wenotedthatit ter the GCN notice time (Yuanet al. 2008). Observations wasonlynecessary tocorrectthedataintheb,uanduvw1 were carried out under fluctuating weather conditions. The bands.Whilstthecorrection toboththeuandbbandswas opticaltransient(OT)wasdetectedbetween25minutesand not large, we found that the uvw1 data required a signifi- 3.5 hours after the trigger. To improve detection signal to cantcorrection, within whichtherewasa largeuncertainty. noise ratio, sets of 4 to 11 images are co-added and expo- Given that there were only two data points from UVOT sures badly affected by weather are excluded. The OT is in this band, we decided to remove them from any further slightly blended with two nearby stars in the ROTSE im- analysis. ages. We therefore subtract the scaled point spread func- tions(PSFs)ofthesetwonearbystarsandthenapplyPSF- matchingphotometryontheOTusingourcustomRPHOT 2.4 Faulkes Telescope North package (Quimby et al. 2006). The analysis is further com- Observations with the Faulkes Telescope North (FTN) plicated by large seeing variation, particularly towards the started at 2008-03-10 09:31:07.3 (UT), 3.188 ks after the endoftheobservation.Thestructureseeninthelightcurve trigger.Datawerereducedinastandardfashion usingiraf (Tody 1986). Calibration was performed using the SDSS datafortheregion(Adelman-McCarthy et al.2007).Forthe 1 SeeLupton(2005)at I and R filters, theSDSSphotometry was converted to the http://www.sdss.org/dr6/algorithms/sdssUBVRITransform.html The early time emission of GRB 080310 5 duringthistimeislikelynotsignificant.TheROTSE-IIIun- thetools provided by the gaia package and with PSF pho- filteredmagnitudesarecalibratedtoSDSSrusingstandard tometrywiththeESO-midas6DAOPHOTcontext(Stetson starsinthepre-burstSDSSobservations(Cool et al. 2008). 1987). The photometric calibration for the NIR was accom- plished by applying average magnitude shifts computed us- 2.6 REM ing 2MASS isolated and non-saturated stars. The optical data were calibrated using instrumental zero points and EarlytimeopticalandNIRdatawerecollectedusingthe60- checkedwithobservationsofstandardstarsinthefieldpro- cm robotic telescope REM (Zerbiet al. 2001; Covino et al. vided by Henden (2008). Linear polarimetry position angle 2004)locatedattheEuropeanSouthernObservatory(ESO) wascorrectedbymeansofobservationsofpolarimetricstan- La Silla observatory (Chile). The telescope simultaneously dard stars in theNGC 2024 region. feeds, by means of a dichroic, the two focal instruments: REMIR (Conconi et al. 2004) a NIR camera, operating in ′ therange1.0 to2.3µm (z ,J,H and K)andROSS(REM Optical Slitless Spectrograph; Tosti et al. 2004) an optical 2.8 WHT imager with spectroscopic (slitless) and photometric capa- Lateimaging was obtained with the4.2m William Herschel bilities(V,R,I).Bothcamerashaveafieldofviewof10× 10 arcmin2. Telescope(WHT),atRoquedelosMuchachosObservatory (LaPalma, Spain)usingtheLong-slit IntermediateResolu- REM reacted automatically after receiving the Swift tionInfraredSpectrograph(LIRIS)initsimagingmode.Ob- alert for GRB 080310, and began observing about 150 sec- servationsconsistedof36×25secondexposuresinH-band, onds after theGRB trigger time (Covino et al. 2008). obtained on the14th March 2008 from 04:36:28 to04:53:41 Optical and NIRdata werereduced following standard UT. The data were reduced following standard procedures procedures. In particular, each single NIR observation with iniraf.Forthephotometriccalibration,weusedstarsfrom REMIR was performed with a dithering sequence of five the2MASS catalogue as reference. images shifted by a few arcseconds. These images are au- tomatically elaborated usingtheproprietary routineAQuA (Testa et al.2004).Thescriptalignstheimagesandco-adds alltheframestoobtainoneaverageimageforeachsequence. 2.9 PAIRITEL Astrometry was performed using USNO-B12 and 2MASS3 catalogue reference stars. PAIRITEL (Bloom et al. 2006) responded to GRB 080310 Photometry was derived by a combination of the SEx- and began taking data at 09:04:58 (UT) in the J, H, tractorpackage(Bertin & Arnouts1996)andthephotomet- andK filterssimultaneously(Perley et al.2008).Theafter- ric tools provided by the gaia4 package. The photometric glow (Chornock et al. 2008) was well-detected in all three calibration fortheNIRwasaccomplished byapplyingaver- filters. Perley et al. (2008) also report on an SED con- age magnitude shifts computed using 2MASS isolated and structed using data from PAIRITEL, KAIT and UVOT non-saturated stars. The optical data were calibrated using (Hoversten & Cummings 2008), allowing a joint fit to be instrumental zero points and checked with observations of made and the estimation of a small amount (AV = 0.10 standard stars in thefield provided by Henden(2008). ±0.05) of SMC-like host-galaxy extinction. 2.7 VLT 2.10 KAIT VLT FORS1 V and R observations for GRB 080310 were automatically activated with the RRM mode5 allowing the The Katzman Automatic Imaging Telescope (KAIT), also at the Lick Observatory (Li et al. 2003), responded to the telescope to react promptly to any alert. The field was ac- trigger and began taking unfiltered exposures starting 42 quired and the observations began less than seven minutes seconds after the trigger time. This paper includes 206 un- after the GRB trigger. Later VLT observations were ob- filtereddatapoints, whichhavebeenreducedin astandard tained with ISAAC at about one day after the burst with way and then calibrated to the R-band (Liet al. 2003). the J, H and K filters. In addition linear polarimetry ob- These data, once calibrated, are shown along with the servationswerecarriedoutwithFORS1withtheV filterat PAIRITELdata in Figure 4. approximately one, two and three daysafter thetrigger. ThefirstKAITdatapointhasacentraltimeof57sec- Optical and NIR data were reduced following stan- onds, with a total exposure time of 30 seconds. Given the dard procedures with the tools of the ESO-eclipse pack- highlyvariablenatureofearlytimeGRBemission,thelarge age (Devillard 1997). Polarimetric data were reduced again error bars on the value, the long duration over which the following standard procedures as discussed in Covino et al. magnitudewasmeasuredand(aslaterdiscussed)itsoutlier (1999,2002,2003).Photometrywasperformed bymeansof nature,wefelt thatthismagnitudedidnotprovideauseful measure the R-band emission over this time. We therefore excluded it from thelater analysis. 2 http://tdc-www.harvard.edu/catalogs/ub1.html 3 http://pegasus.phast.umass.edu/ 4 http://star-www.dur.ac.uk/pdraper/gaia/gaia.html 5 http://www.eso.org/sci/observing/phase2/SMSpecial /RRMObservation.html 6 http://www.eso.org/sci/software/esomidas/ 6 O.M. Littlejohns et al. Table 1. ESO-VLT polarimetric observations of the late after- glowofGRB080310 t−t0 (s) Texp (s) Polarization(%) Band 87171 1447 <2.5 V 169501 1447 <2.5 V 253724 4607 <2.6 V the physical beaming of the outflow (Ghisellini & Lazzati 1999; Sari 1999). Therefore the polarization time-evolution isinprincipleapowerfuldiagnosticoftheafterglowphysics, and many attempts were carried out to compare observa- tionstomodels(e.g.,Lazzati et al.2003,2004)withpartic- ular emphasis to thejet structure (Rossi et al. 2004). During the polarimetric observations of GRB 080310 theafterglowshowedasmoothdecay(seeFigure3)without anydetectabletemporalbreakorspectralchange.Polariza- tion below ∼ 2.5% cannot put specific constraints on the afterglow modeling or the jet structure. These results, are however compatible with what would be expected at late times, as this is when the forward shock should dominate Figure 4. PAIRITEL and KAIT light curves for GRB 080310. TheKAITunfiltereddataarecalibratedtotheR-band.Alldata emission and the magnetization signal of the fireball is lost havebeencorrectedfordustextinction. in the interaction with thesurrounding medium. 2.11 Gemini 2.13 Observations from literature OurlastopticaldatawereacquiredwithGemini-Northusing DatafromthePalomar60inchtelescope(Cenko et al.2006) the Gemini Multi-Object Spectrograph (GMOS) in imag- were obtained from the Palomar 60 inch-Swift Early Opti- ing mode with ther-band filter. The observations began at cal Afterglow Catalog (Cenkoet al. 2009), in which the 29 st st 10:22UTon19thMarch2008andconsistedof5×150second GRBsbetween the1 of April2005 and the31 of March exposures. The data were reduced using the gemini-gmos 2008 with P60 observations beginning within the first hour routineswithiniraf.Nosignificantfluxwasdetectedatthe aftertheinitialSwift-BATtriggerarepresented.Cenko et al. location of theafterglow, as reported in Table A5. (2009) reduce data in the iraf environment, using a cus- tom pipeline detailed in Cenko et al. (2006). Magnitudes were calculated using aperture photometry and calibration 2.12 Polarization performed using the USNO-B1 catalog7 and the data were corrected for dust extinction using the extinction maps of Three linear polarimetric observation sets were carried out (Schlegel et al. 1998). with the ESO-VLT during the late afterglow evolution as Further published data for GRB 080310 were obtained shown in Table1. The observations allowed us to derive from Kann et al. (2010), in which SMARTS-ANDICAM ratherstringent upperlimits although stillcompatible with data are detailed as part of an extensive survey of opti- past late-time afterglow detections (Covino et al. 2005, see cal data for GRBs in both the pre-Swift and Swift eras. also, however, Bersier et al. 2003) and substantially lower Kann et al. (2010) reduce their data using standard pro- than the early time afterglow measurement by Steele et al. cedures in iraf and midas. Both aperture and PSF pho- (2009) for GRB 090102. In the case of GRB 090102, how- tometry were used in the derivation of magnitudes, when ever, the detection is taken at 160 seconds; a time when comparing to standard calibrator stars. Steele et al.(2009)arguethatthefluxshouldbedominated The 0.6m Super-LOTIS (Livermore Optical Tran- by reverseshock emission. sient Imaging System) telescope, located at the Stew- The detection of linear polarization at the level of a ard Observatory (Kitt Peak, Arizona; P´erez-Ram´ırez et al. few percent in the light from GRB optical afterglow is 2004) began R-band observations of the error region of well within the prediction of the external shock scenario GRB 080310 at 08:38:43 UT, 44 seconds after the start of (Zhang& M´esz´aros 2004, and references therein) and in- the burst (Milne & Williams 2008). The OT detected by deeditisstilloneofthemostrelevantobservationalfindings Chornock et al. (2008) and confirmed by Cummings et al. supportingit(e.g.,Covino2010).Ontheotherhand,acom- (2008) was not apparent in the initial images, even when prehensiveframeworkpredictingthepolarizationdegreeand stacking the first three 10 second exposures. However, the position angle evolution during the afterglow had to deal subsequent 20 second exposures do show the optical tran- with the increasing complexity and variety of behaviours sientwithoutstacking,whichsuggeststhattheGRBbright- shown by the afterglow population. In general, the late- ened in the R-band during the first two minutes after de- afterglowopticalpolarization isrelatedtothreemainingre- dients:theemissionprocessabletogeneratehighlypolarized photons(i.e.,synchrotron),theultra-relativisticmotionand 7 http://www.nofs.navy.mil/data/fchpix The early time emission of GRB 080310 7 tection. A nearby USNO-B star was used to derive the R Table 2. Properties of the prompt pulses identified in the BAT magnitude. and XRT light curves. These include the peak time (Tpk), peak energy at this time(Epk), the risetime(Tr) and arrivaltimeof thelastphoton(Tf)foreachpulse. 3 MODELLING Pulse Tpk (s) Epk (keV) Tr (s) Tf (s) b1 To model the emission of GRB 080310 we begin by fitting 1 -52.8 200 9.7 44.5 -1.49 the BAT and XRT data, before then extending the model 2 -16.0 200 5.0 6.0 -1.20 into thelower energy bands. 3 -4.6 200 3.8 11.7 -0.40 4 1.8 200 2.7 17.2 -1.50 5 159.0 12.3 25.3 74.0 -0.30 3.1 Initial modelling of the high energy emission 6 191.6 13.4 12.4 39.8 -0.02 7 210.0 21.3 10.0 46.7 -0.16 We expand on the previous work done by Willingale et al. 8 235.0 58.0 8.0 24.4 -0.13 (2010), where a sample of 12 GRBs were selected and fit- 9 251.8 42.0 10.0 52.4 -0.20 tedusingthepulsemodelofGenet & Granot(2009).Inthis 10 282.0 15.8 10.0 40.0 -0.10 model, thepromptemission componentof GRBemission is 11 308.7 16.1 15.0 32.5 0.24 split into a series of pulses, where each pulse is considered 12 342.7 7.8 15.0 40.1 -0.18 13 366.0 34.3 10.0 11.5 -0.38 tobetheresultofarelativistically expandingthinspherical 14 390.0 1.2 23.0 56.1 0.21 shell that emits isotropically. It was assumed that each of 15 513.8 3.8 31.1 193.8 -1.26 the pulses was in the fast cooling regime (Sariet al. 1998) 16 582.1 2.4 43.9 66.5 -0.09 and that each of the X-ray and γ-ray spectra could be fit- ted with a temporally evolving Band function (Band et al. 1993). A Band spectral energy distribution is a smoothly finally αa is the index that governs the temporal decay of broken power-law. Below, we show a modified Band func- thepower-law phase. tion, which includes temporal evolution. Combiningthepromptpulsesandtheafterglowcompo- nent we adopted the same method of fitting the data from zb1−1e−z, z ≤ b −b , B(z)=B(cid:26) zb2−1(b1−b2)b1−b2e−(b1−b2), z > b11−b22, (1) t(h20e10S)w,ibftyBfiArsTtidaenndtifXyRinTgtihnestirnudmiveindtusalaspuWlseilsliinngathleeeBtAaTl. lightcurveandallowingtheirparameterstobefittedbymin- where imizingtheχ2 fitstatistic forboththeBATandXRTlight d z= E T −Tej . (2) curves.Theafterglowcomponentwasthenfittedbyallowing (cid:18)Ef(cid:19)(cid:18) Tf (cid:19) theroutinetofindtheoptimumvaluesforthecharacteristic times and normalizing fluxshown in Eqn 3. The times in Eqn 2 are all in the observed frame, where The simultaneous BAT and XRT fit is plotted in Fig- Tej is thetime of shell ejection from thecentral engine and ure 5 and shows several important characteristics of both T is the time at which the last photon arrives from the f the burst and the model. Firstly, there is a lot of structure shell along the line of sight. B is the normalisation to the evidentinthelightcurves,particularlyduringthefirst1000 Bandfunction,withb andb beingthelowandhighenergy 1 2 seconds of the XRT light curves. In this fit, sixteen unique photonindicesoftheBandfunction,respectively.Thespec- pulses have been identified and can be seen to fit the XRT tral shape of each pulse evolves in time as determined by data accurately. In the original fit from Willingale et al. thetemporalindexd.InWillingale et al. (2010)d=-1due (2010) there were only ten pulses, however this failed to to the assumption that the emission process is synchrotron fully model some of the structure in the softest BAT band under the standard internal shock model. It is this value (15-25 keV) between 200 and 400 seconds. The additional of temporal index that was also used by Genet & Granot pulses now provide a better fit to this time, across all the (2009)intheoriginal derivationofthepulseprofiles,mean- BATandXRTbandswith areducedχ2 statisticof1.68for ingthatonlyavalueofd=-1isstrictlyconsistentwiththe 1212 degrees of freedom. The quality of this fit is formally original pulse model. notstatisticallyacceptable,howeverthelargestcontribution The prompt pulse model describes the morphology of to the χ2 value is from small scale intrinsic fluctuations in the prompt light curve and the rapid decay phase as ob- the data. The properties of the sixteen pulses are listed in served in the high-energy bands. In addition to this, an af- Table 2. terglow component was also included in the modelling as outlined in Willingale et al. (2007). The afterglow compo- nenthasafunctionalformasoutlinedinEqn3,whichcom- 3.2 Extrapolating to the optical and IR prisesanexponentialphasethattransitionsintoapower-law Previously, no optical or IR data have been included when decay. modellingthelightcurveandafittoonlytheSwiftBATand fa(t)=(cid:26) FFaaexTtpa(cid:0)α−aαa−extTαpaa(cid:1)−etTxrp(cid:0),−tTr(cid:1), tt <≥ TTaa., (3) XrfoaRrpTiGdRdraeBtsap0oh8n0as3se1b0oefewnoeppterixoctadelunacdenddth(IeFRimgiuonrdseterl5u)tm.oHeinnotwcsleutvdoeert,thgheievsteernigntgehewer (cid:0) (cid:1) (cid:0) (cid:1) In Eqn 3, fa(t) is the flux from the afterglow at time t, Fa sources of data. givesthefluxatthetransitiontimebetweentheexponential ThesimplestapproachtofittingtheopticalandIRlight and the power-law components, Ta. Tr is the rise time and curves for GRB 080310 was to use the fitting routine from 8 O.M. Littlejohns et al. Figure5.BATandXRTdata,showingthefittedWillingaleetal.(2010)model.Leftpanels:BATbands(Toptobottom:100-350 keV, 50-100 keV,25-50 keV&15-25 keV).Rightpanels:XRTbands(Top:XRThardband,bottom:XRTsoftband).Pulseparametershave been fixed, whilst the afterglow parameters have been fitted. Solid lines show the total fit (pulses and afterglow), whilst dashed lines showonlytheafterglowcomponent tothefit. Willingale et al. (2010) and simply extrapolate the Band tween the X-ray and the optical and IR energies. Such a functionsforboththepulsesandalsotheafterglowtothese break is expected in a synchrotron spectrum, but can also lower energies. In this initial attempt all of the parameters beenseeninthermalspectraastheRayleigh-Jeanstail.Ad- previouslydiscussedwereheldatthevaluesobtainedforthe ditionally,theafterglow prediction from theBATandXRT highenergyfit,toseewhatmodificationsmightbenecessary fitisnotconsistentwiththeopticaldataatlatetimeswhere to both components. thepromptcomponenttothelightcurveisnegligible.Figure Whilst it provides an acceptable fit to the XRT and 6alsoshowsthatthetemporaldecayindexofthepower-law BAT bands, the original pulse model vastly over predicts phase of the afterglow gives rise to a decay which is more the optical and IR fluxes from the pulses (Figure 6), which rapid than theoptical data suggest. implies there must be a break in the pulse spectrum be- Analternativemethod ofreducingopticalfluxistoin- The early time emission of GRB 080310 9 Figure 6. Optical and IR light curves obtained by extending the high energy fit to the prompt pulses and afterglow component.The spectralshapeofthepulsesisassumedtobeaBandfunctionextrapolatedbacktotheopticalandIRdata(K,H,J,I,R,V,B andU bands).Verticaldashedlinesdenote timesofinterestlaterdiscussed,atwhichtheSEDsareconsideredatinmoredetail.Thecoloured dashedlinesshowtheafterglowcomponent totheextrapolation foreachband. voke dust absorption, however, as reported in Perley et al. the pulse is at a maximum. The entire pulse spectrum al- (2008), extinction duetodust is low at AV = 0.10 ±0.05 ready evolves in time, according to the index d as shown at an average time of T (+1,750) seconds for GRB 080310. in Eqn 2, and so the expectation was that the time evolu- 0 As not only the late time emission seems unaffected by tionofEawouldberelated.Previouslyavalueofd=-1was such absorption, butalso by 2,000 seconds after trigger, we used,whichwhenintegratingoftheEqualArrivalTimeSur- chose to favour a low energy spectral break in our mod- face (EATS) means a pure Band function can be recovered elling of the optical emission. The following sections of this in the source frame given a Band function in the observed paper detail the implementation of such alterations to the frame. Other values of d prevent assumptions to be made Willingale et al. (2010) model. concerningthesourcespectrum.Weallowedtheindexda to befittedindependently,totestwhethertheevolutionofthe break was the same as the rest of the spectrum. Again, we 3.3 Modifications to the model note that only da = -1 is fully consistent with the original Genet & Granot(2009)pulsemodel.Thefunctionalformof 3.3.1 An additional spectral break thenew spectral model for each pulse is shown in Eqn 4. To reduce the flux from each pulse in the optical and IR light curves of GRB 080310 we introduced an additional AEpa−1exp −EEca Ea−α, E ≤ Ea, break to the spectrum for each prompt pulse. An example N(E)= AEb1−1exp(cid:0)−EEc (cid:1), Ea < E ≤ Epk, (4) of such a spectrum is shown in Figure 7, where a regular  AEb2−1exp((cid:0)α−(cid:1)β)E−(α−β), E > E . pk pk Band function has a low energy break below E . To fully pk  describethisbreak,threeparameters areneeded;avalueof The model spectrum for photon emission assumes several the break energy (Ea), the spectral index of the power-law things. Firstly, there is a single underlying population of slope in the low energy regime (pa) and a temporal index relativistic electrons, whose energies can be described by a which describes how the break energy evolves in time (da). broken power law such as that described in Shen & Zhang The value of Ea is defined at Tpk, when the emission from (2009).Suchaspectrumofelectronenergiesleadstoasimi- 10 O.M. Littlejohns et al. nels.Thechangein αa requiredwas foundtobesmall, at a valueofapproximately0.2,buttheassociatederrorsineach instanceshowedittobeinconsistentwithzero.Thismaybe explained physically by a slight curvature in the spectrum oftheafterglow betweentheopticalandX-raybands.This, however, is not a large enough difference to suggest a spec- tralbreak,suchastheoneintroducedtothepromptpulses. The GRB afterglow flux is also thought to be synchrotron emission. However, at these late times the radiation is usu- allyassumed tocomefrom anoptically thinplasma,andso aself-absorption break would beexpectedat energies lower than the optical bands observed for GRB 080310. We also ruledoutcontaminationintheopticalandIRwavebandsas being thesource of this difference as any emission from the Figure7.Examplemodifiedspectrumincludingadditionalspec- host should be constant with time. This should add a con- tralbreak. stant offset to the data, and should the afterglow emission reach a comparable order of magnitude, a plateau at the larphoton spectrum:asingly brokenpower-law, whichcor- endof theoptical decay would beexpected.Thisplateau is responds to theobserved Band function. However,it would not observed, and the difference between the low and high be unphysical for this spectral shape to extend indefinitely energytemporaldecayindicesoftheafterglowisalsodeter- tolowenergies,particularlyastheelectronenergyspectrum minedbydatapriortowhenthehostwouldbeseentomake hasanassociatedminimumenergyEm.Thisminimumelec- a significant impact on theoptical and IR light curves. tronenergyhasarelatedemissionenergyatafrequencyνm. With no prompt component capable of rising quickly, With no self-absorption, the spectral index to the emission we also had to modify the manner in which the afterglow spectrumchangesto 13 atphotonfrequenciesbelowνm,but rises, to account for the rapid increase in flux seen in the if self-absorption is present then at low energies we would V-band data at approximately 100 seconds in Figure 3. To expect to see a steeper spectral index of 2 if the absorption do this, we introduced a third part to the functional form frequency is less than νm or alternatively a spectral index that is shown Eqn 3. This extra regime is shown in thetop of 52 in the case of νm < νa. In the former case, if there is line of Eqn 5, and describes a power-law rise in flux, with aninterveningspectralrangebetweenνa andνm,aspectral anindexof2.Wealso allowed theafterglow tobelaunched index of 1 is expected between these two frequencies. at a time that was independent to the trigger time of the 3 To keep themodel simple, dueto thelimited natureof GRB, by introducing a launch time (T), which offsets the l thespectralinformation available,weonlyincludedasingle afterglowintimetobetterfitthetimingoftheriseobserved additionalbreakinthespectrum.Byallowing theresultant intheV-band.ThesemodificationsareshowninEqn5and spectralindextobeconstrainedbytheobservations,rather areusedfortheafterglow-dominated modelling only.Inthe than assuming a value, we hope to better understand the caseofthepromptdominatedfit,wereturntotheafterglow physics required to explain the spectrum of prompt GRB model of Willingale et al. (2007) as explained in §3.5. pulses.  Faexp(cid:16)αa− (TrT−a−TlT)lαa(cid:17)(cid:16)Ttr−−TTll(cid:17)2, t ≤ Tr, 3Abti.rcs3ea.asl2keaesnniTdgihnnIeiRfiFaclfaiitggneuhtrrlgtyelocow6uv,revprerspormperdpioitcrtpttuohlse1e0so,b0w0sei0trhvsoeeducotflnaudxsl.oiwBnytehinenetrorgpoy-- fa(t)= FFaae(cid:16)xTpta−(cid:16)−TαTlla(cid:17)−−α(atT−,aT−l)Tαla(cid:17), Ttr><Tat.≤ Ta, (5) ducing a low energy break to the pulse spectra, we hoped 3.4 Afterglow dominated fit to completely remove the prompt contribution to the total emission, and model the entire light curve with afterglow Giventhesmooth natureoftheopticalandIRlightcurves, emission. One reason for considering this is the smooth na- ourinitialuseofthelow-energyspectralbreakintheprompt ture of the optical light curves. Whilst there is some varia- pulsespectrumwastoremovethepromptcomponenttothe tionduringtheplateauseenbetween200and3,000seconds, optical light curves entirely, and try and explain the early thisisatalowlevel,andhappenssmoothlyoveralargepe- timebehaviourwithonlyafterglowemission.Todothis,the riodoftime.Ifthepromptpulsesobservedathigherenergies prompt pulses were initially switched off from the optical werealsothedominantsourceofemissionattheseenergies, andnearinfraredfit.Withtheseremoved,theafterglowpa- thenwe might expectto seesimilar structurein theoptical rameters Tl, Tr,Ta, Fa, αa and the change between optical and IR light curvesto that shown in Figure 5. and high energy values of αa were fitted. Having obtained The first issue to address with the afterglow fit was valuesfor theseparameters using thefittingroutine, we re- to account for the small difference in temporal decay in- introduced the prompt pulses, allowing pa to be fitted, to dex (αa) between the higher and lower energy bands. To find the minimum break in the pulse spectra required to do so, we included an additional parameter that described removethepromptopticalandIRflux.Wherethefitstatis- the difference in this index, and allowed it to be fitted. By tic tended to unphysical values of pa we checked the effects including this, we could account for the slightly shallower of forcing pa to be 25. When this was undertaken,we found decay of the power-law phase in the optical and IR chan- thatwhilsttherewasanincreaseinthereducedχ2 statistic,