Detection of a supernova signature associated with GRB 011121
a r X i v :a s t r o -p h /0203391v 2 28 M a y 2002
Published in the Astrophysical Journal Letters ,vol.572,L45–L49
Preprint typeset using L A T E X style emulateapj v.14/09/00
DETECTION OF A SUPERNOVA SIGNATURE ASSOCIATED WITH GRB 0111211
J.S.Bloom 2,S.R.Kulkarni 2,P.A.Price 2,3,D.Reichart 2,T.J.Galama 2,B.P.
Schmidt 3,D.A.Frail 2,4,E.Berger 2,P.J.McCarthy 9,R.A.Chevalier 5,J.C.Wheeler 6,J.P.Halpern 7,D.W.Fox 2,S.G.Djorgovski 2,F.A.Harrison 2,R.Sari 8,T.S.Axelrod 3,R.A.Kimble 10,J.Holtzman 11,K.Hurley 12,F.Frontera 13,14,L.Piro 15,&E.Costa 15
Received 2002March 21;accepted 2002May 2;published 2002May 20
ABSTRACT
Using observations from an extensive monitoring campaign with the Hubble Space Telescope we present the detection of an intermediate time ?ux excess which is redder in color relative to the afterglow of GRB 011121,currently distinguished as the gamma-ray burst with the lowest known redshift.The red “bump,”which exhibits a spectral roll-over at ～7200?A ,is well described by a redshifted Type Ic supernova that occurred approximately at the same time as the gamma-ray burst event.The inferred luminosity is about half that of the bright supernova 1998bw.These results serve as compelling evidence for a massive star origin of long-duration gamma-ray bursts.Models that posit a supernova explosion weeks to months preceding the gamma-ray burst event are excluded by these observations.Finally,we discuss the relationship between spherical core-collapse supernovae and gamma-ray bursts.
Subject headings:supernovae:general —gamma rays:bursts —supernovae:individual (SN 1998bw)
1.introduction
Two broad classes of long-duration gamma-ray burst (GRB)progenitors have survived scrutiny in the afterglow era:the coalescence of compact binaries (see Fryer et al.1999for review)and massive stars (Woosley 1993).More exotic explanations (e.g.,Paczy′n ski 1988;Carter 1992;Dermer 1996)fail to reproduce the observed redshift dis-tribution,detection of transient X-ray lines,and/or the distribution of GRBs about host galaxies.
In the latter viable scenario,the so-called “collap-sar”model (Woosley 1993;MacFadyen &Woosley 1999;Hansen 1999),the core of a massive star collapses to a com-pact stellar object (such as a black hole or magnetar)which then powers the GRB while the rest of the star explodes.We expect to see two unique signatures in this scenario:a rich circumburst medium fed by the mass-loss wind of the progenitor (Chevalier &Li 1999)and an underlying supernova (SN).Despite extensive broadband modeling of afterglows,unambiguous signatures for a wind-strati?ed circumburst media have not been seen (e.g.,Frail et al.2000;Berger et al.2001).
There has,however,been been tantalizing evidence for an underlying SN.The ?rst association of a cosmologi-cally distant GRB with the death of a massive star was found for GRB 980326,where a clear excess of emission was observed,over and above the rapidly decaying after-glow component.This late-time “bump”was interpreted as arising from an underlying SN (Bloom et al.1999)since,unlike the afterglow,the bump was very red.GRB 970228,also with an intermediate-time bump and characteristic SN spectral rollover,is another good candidate (Reichart 1999;Galama et al.2000).
Suggestions of intermediate-time bumps in GRB light curves have since been put forth for a number of other GRBs (Lazzati et al.2001;Sahu et al.2000;Fruchter et al.2000;Bj¨o rnsson et al.2001;Castro-Tirado et al.2001;Sokolov 2001;Dar &R′u jula 2002).Most of these results are tentative or suspect with the SN inferences re-lying on a few mildly deviant photometric points in the afterglow light curve.Even if some of the bumps are real,a number of other explanations for the physical origin of such bump have been advanced:for example,dust echoes
1
Based on observations with the NASA/ESA Hubble Space Telescope,obtained at the Space Telescope Science Institute,which is operated by the Association of Universities for Research in Astronomy,Inc.under NASA contract No.NAS5-26555.
2Division of Physics,Mathematics and Astronomy,105-24,California Institute of Technology,Pasadena,CA 91125
3Research School of Astronomy &Astrophysics,Mount Stromlo Observatory,via Cotter Rd.,Weston Creek 2611,Australia 4National Radio Astronomy Observatory,Socorro,NM 87801
5Department of Astronomy,University of Virginia,P.O.Box 3818,Charlottesville,VA 22903-08186Astronomy Department,University of Texas,Austin,TX 78712
7Columbia Astrophysics Laboratory,Columbia University,550West 120th Street,New York,NY 100278Theoretical Astrophysics 130-33,California Institute of Technology,Pasadena,CA 911259Carnegie Observatories,813Santa Barbara Street,Pasadena,CA 91101
10Laboratory for Astronomy and Solar Physics,NASA Goddard Space Flight Center,Code 681,Greenbelt,MD 2077111Department of Astronomy,MSC 4500,New Mexico State University,P.O.Box 30001,Las Cruces,NM 8800312University of California at Berkeley,Space Sciences Laboratory,Berkeley,CA 94720-7450
13Istituto Astro?sica Spaziale and Fisica Cosmica,C.N.R.,Via Gobetti,101,40129Bologna,Italy 14Physics Department,University of Ferrara,Via Paradiso,12,44100Ferrara,Italy
15Istituto Astro?sica Spaziale,C.N.R.,Area di Tor Vergata,Via Fosso del Cavaliere 100,00133Roma,Italy
1
2
(Esin&Blandford2000;Reichart2001),shock interaction with circumburst density discontinuities(e.g.,Ramirez-Ruiz et al.2001),and thermal re-emission of the afterglow light(Waxman&Draine2000).To de?nitively distinguish between the SN hypothesis and these alternatives,detailed spectroscopic and multi-color light curve observations of intermediate-time bumps are required.
It is against this background that we initiated a pro-gram with the Hubble Space Telescope(HST)to sample afterglow light curves at intermediate and late-times.The principal attractions of HST are the photometric stabil-ity and high angular resolution.These are essential in separating the afterglow from the host galaxy and in re-constructing afterglow colors.
On theoretical grounds,if the collapsar picture is true, then we expect to see a Type Ib/Ic SN(Woosley1993). In the?rst month,core-collapsed supernova spectra are essentially characterized by a blackbody(with a spectral peak near～5000?A)modi?ed by broad metal-line absorp-tion and a strong?ux suppression blueward of～4000?A in the restframe.For GRBs with low redshifts,z～<1,the e?ect of this blue absorption blanketing is a source with an apparent red spectrum at observer-frame optical wave-lengths;at higher redshifts,any supernova signature is highly suppressed.For low redshift GRBs,intermediate-time follow-up are,then,amenable to observations with the Wide Field Planetary Camera2(WFPC2).In this Letter we report on WFPC2multi-color photometry of GRB011121(z=0.36;Infante et al.2001)and elsewhere we report on observations of GRB010921(z=0.451;Price et al.2002a).In a companion paper(Price et al.2002b; hereafter Paper II),we report a multi-wavelength(radio, optical and NIR)modeling of the afterglow.
2.observations and reductions
2.1.Detection of GRB011121and the afterglow
On2001November21.7828UT,the bright GRB011121 was detected and localized by BeppoSAX to a5-arcmin radius uncertainty(Piro et al.2001).Subsequent obser-vations of the error circle re?ned by the IPN and Bep-poSAX(see Paper II)revealed a fading optical transient (OT)(Wyrzykowski et al.2001;Stanek et al.2001).Spec-troscopic observations with the Magellan6.5-m telescope revealed redshifted emission lines at the OT position(z= 0.36),indicative of a bright,star-forming host galaxy of GRB011121(Infante et al.2001).
2.2.HST Observations and reductions
For all the HST visits,the OT and its underlying host were placed near the serial readout register of WF chip 3(position WF ALL)to minimize the e?ect of charge transfer(in)e?ciency(CTE).The data were pre-processed with the best bias,dark,and?at-?eld calibrations avail-able at the time of retrieval from the archive(“on–the–?y”calibration).We combined all of the images in each ?lter,dithered by sub-pixel o?sets,using the standard IRAF/DITHER2package to remove cosmic rays and pro-duce a better sampled?nal image in each?lter.An image of the region surrounding the transient is shown in?gure 1.The point source was detected at better than20σin epochs one,two and three in all?lters,and better than5σin epoch four.
N
E
Host
5"
Fig. 1.—Hubble Space Telescope image of the?eld of GRB 011121on2001December4–6UT.This false-color image was constructed by registering the?nal drizzled images in the F555W (blue),F702W(green)and F814W(red)?lters.The optical tran-sient(OT)is clearly resolved from the host galaxy and resides in the outskirts of the morphologically smooth host galaxy.Follow-ing the astrometric methodology outlined in Bloom et al.(2002), we?nd that the transient is o?set from the host galaxy(883±7) mas west,(86±13)mas north.The projected o?set is(4.805±
0.035)kpc,almost exactly at the host half-light radius.Sources
“A”and“B”are non-variable point sources that appear more red than the OT and are thus probably foreground stars.
Given the proximity of the OT to its host galaxy, the?nal HST images were photometered using the IRAF/DAOPHOT package which implements PSF-?tting photometry on point-sources(Stetson1987).The PSF local to the OT was modeled with PSTSELECT and PSF using at least15isolated stars detected in the WF chip3with an adaptive kernel to account for PSF variations across the image(VARORDER=1).The re-sulting photometry,reported in Table1,was obtained by ?nding the?ux in an0′′.5radius using a PSF?t.We cor-rected the observed countrate using the formulation for CTE correction in Dolphin(2000)with the most up-to-date parameters16;such corrections,computed for each individual exposure,were never larger than8%(typically 4%)for a?nal drizzled image.We estimated the uncer-tainty in the CTE correction,which is dependent upon source?ux,sky background,and chip position,by com-puting the scatter in the CTE corrections for each of the images that were used to produce the?nal image.The magnitudes reported in the standard bandpass?lters in Table1were found using the Dolphin prescription.
3.results
In?gure2we plot the measured?uxes from our four HST epochs in the F555W,F702W,F814W and F850LP ?lters.We also plot measurements made at earlier times
16See https://www.wendangku.net/doc/333212620.html,/staff/dolphin/wfpc2
3
(0.5d Corrections for color e?ects between the ground-based ?lters and HST?lters were taken to be negligible for the purpose of this exercise. The estimated contribution from the afterglow is heav-ily weighted by the available data:our ground-based data (and those reported in the literature so far)are primarily at early times.Roughly,over the?rst week,the afterglow exhibits a simple power law decay.The afterglow contri-bution derived from our NIR data and optical data from the literature(see Paper II)is shown by the dashed line in each panel.No afterglow light curve breaks(e.g.,from jetting)were assumed. Garnavich et al.(2002)drew attention to an excess of ?ux(in R-band),at a time13days after the GRB,with respect to that expected from the power-law extrapola-tion of early-time afterglow emission;they suggested the excess to arise from an underlying SN.As can vividly be seen from our multi-color data,the excess is seen in all bands and over several epochs. We used the light curve and spectra17of the well-studied Type Ic supernova SN1998bw(Galama et al.1998; McKenzie&Schaefer1999)to create a comparison tem-plate broad-band light curve of a Type Ic supernova at redshift z=0.36.Speci?cally,the spectra of SN1998bw were used to compute the K-corrections between observed photometric bands of1998bw and HST bandpasses(fol-lowing Kim et al.1996and Schmidt et al.1998).A?atΛcosmology with H0=65km s?1Mpc?1and?M=0.3was assumed and we took the Galactic foreground extinction to SN1998bw of A V=0.19mag(Galama et al.1998). Since dimmer Ic SNe tend to peak earlier and decay more quickly(see?g.1of Iwamoto et al.1998),much in the same way that SN Ia do,we coupled the?ux scaling of SN1998bw with time scaling in a method analogous to the“stretch”method for SN Ia distances(Perlmutter et al.1997).To do so,we?t an empirical relation be-tween1998bw and1994I to determine the?ux-time scal-ing.We estimate that a1998bw-like SN that is dimmed by55%(see below),would peak and decay about17% faster than1998bw itself.Some deviations from our sim-ple one-parameter template are apparent,particularly in the F555W band and at late-times. In?gure3,we plot the spectral?ux distributions(SFDs) of the intermediate-time bump at the four HST epochs.A clear turn-over in the spectra in the?rst3epochs is seen at about7200?A.The solid curve is the SFD of SN1998bw transformed as described above with the associated2-σerrors.Bearing in mind that there are large systematic uncertainties in the template(i.e.,the relative distance moduli between SN1998bw and GRB011121)and in the re-construction of the red bump itself(i.e.,the Galactic extinction toward GRB011121and the contribution from the afterglow in the early epochs),the consistency between the measurements and the SN is reasonable.We consider the di?erences,particularly the bluer bands in epoch one, to be relatively minor compared with the overall agree-ment.This statement is made in light of the large observed spectral diversity of Type Ib/Ic SNe(see,for example,?g-ure1of Mazzali et al.2002). 4.discussion and conclusions We have presented unambiguous evidence for a red, transient excess above the extrapolated light curve of the afterglow of GRB011121.We suggest that the light curve and spectral?ux distribution of this excess appears to be well represented by a bright SN.While we have not yet explicitly compared the observations to the expectations of alternative suggestions for the source of emission(dust echoes,thermal re-emission from dust,etc.),the simplic-ity of the SN interpretation—requiring only a(physically motivated)adjustment in brightness—is a compelling(i.e., Occam’s Razor)argument to accept our hypothesis.Given that the red bump detections in a number of other GRBs occur on a similar timescale as in GRB011121,any model for these red bumps should have a natural timescale for peak of～20(1+z)day;in our opinion,the other known possibilities do not have such a natural timescale as com-pared with the SN hypothesis.Indeed,if our SN hypothe-sis is correct,then the?ux should decline as an exponen-tial from epoch four onward.The ultimate con?rmation of the supernova hypothesis is a spectrum which should show characteristic broad metal-line absorption of the ex-panding ejecta(from,e.g.,Ca II,Ti II,Fe II). We used a simplistic empirical brightness–time stretch relation to transform1998bw,showing good agreement be-tween the observations and the data.If we neglect the time-stretching and only dim the1998bw template,then the data also appear to match the template reasonably well,however,the discrepancies in the bluer bands be-come somewhat larger and the?ux ratios between epochs are slightly more mismatched.The agreement improves if we shift the time of the supernova to be about～3–5 days(restframe)before the GRB time.Occurrence times more than about ten days(restframe)before the GRB can be ruled out.This observation,then,excludes the origi-nal“supranova”idea(Vietri&Stella1998),that posited a supernova would precede a GRB by several years(see eq.[1]of Vietri&Stella1998).Modi?ed supranova sce-narios that would allow for any time delay between the GRB and the accompanying SN,albeit ad hoc,are still consistent with the data presented herein18. Regardless of the timing between the SN explosion and the GRB event(constrained to be less than about10days apart),the bigger picture we advocate is that GRB011121 resulted from an explosive death of a massive star.This conclusion is independently supported by the inference, from afterglow observations of GRB011121(Paper II),of a wind-strati?ed circumburst medium. The next phase of inquiry is to understand the details of the explosion and also to pin down the progenitor pop-ulation.A large diversity in any accompanying SN com-ponent of GRBs is expected from both a consideration of 17Spectra were obtained through the Online Supernova Spectrum Archive(SUSPECT)at https://www.wendangku.net/doc/333212620.html,/～suspect/index.html. 18The explosion date of even very well-studied supernovae,such as1998bw,cannot be determined via light curves to better than about3 days(e.g.,Iwamoto et al.1998).This implies that future photometric studies might not be equipped to distinguish between contemporaneous SN/GRB events and small delay scenarios. 4 0.1 1 10 100 F l u x d e n s i t y (μJ y ) F555W F702W 1 10 100 t (day) 0.1 1 10 100 F l u x d e n s i t y (μJ y ) F814W 1 10 100 t (day) F850LP Fig. 2.—Light-curves of the afterglow and the intermediate-time red bump of GRB 011121.The triangles are our HST photometry in the F555W,F702W,F814W and F850LP ?lters (all corrected for the estimated contribution from the host galaxy),and the diamonds are ground-based measurements from the literature (Olsen et al.2001;Stanek &Wyrzykowski 2001).The dashed line is our ?t to the optical afterglow (see Paper II),the dotted line is the expected ?ux from the template SN at the redshift of GRB 011121,with foreground extinction applied and dimmed by 55%to approximately ?t the data,and the solid line is the sum of the afterglow and SN components.Corrections for color e?ects between the ground-based ?lters and the HST ?lters were taken to be negligible for the purpose of this exercise. SNe themselves and the explosion mechanism.The three main physical parameters of a Type Ib/Ic SN are the total explosive energy,the mass of the ejecta,and the amount of Nickel synthesized by the explosion (M Ni ).The peak lu-minosity and time to peak are roughly determined by the ?rst two whereas the exponential tail is related to M Ni .Ordinary Ib/Ic SNe appear to show a wide dispersion in the peak luminosity (Iwamoto et al.1998).There is lit-tle ab initio understanding of this diversity (other than shifting the blame to dispersion in the three parameters discussed above). It is now generally accepted that GRBs are not spherical explosions and are,as such,usually modeled as a jetted out?ow.Frail et al.(2001)model the afterglow of GRBs and have presented a compilation of opening angles,θ,ranging from less than a degree to 30degrees and a median of 4degrees.If GRBs have such strong collimation then it is not reasonable to assume that the explosion,which explodes the star,will be spherical.We must be prepared to accept that the SN explosion is extremely asymmetric and thus even a richer diversity in the light curves.This expected diversity may account for both the scale factor di?erence between the SN component seen here and in SN 1998bw seen in ?gure 3.Indeed,there has been a signi?- cant discussion in the literature as to the degree which the central engine in GRBs will a?ect the overall explosion of the star (Woosley 1993;Khokhlov et al.1999;MacFadyen &Woosley 1999;H¨o ?ich et al.1999;MacFadyen et al.2001).These models currently have focused primarily on the hydrodynamics and lack the radiative modeling neces-sary to compare observations to the models. Clearly,the next step is to obtain spectroscopy (and perhaps even spectropolarimetry)and to use observations to obtain a rough measure of the three-dimensional veloc-ity ?eld and geometry of the debris.As shown by GRB 011121the SN component is bright enough to undertake observations with the largest ground-based telescopes.We end by noting the following curious point.The to-tal energy yield of a GRB is usually estimated from the gamma-ray ?uence and an estimate of θ(see Frail et al.2001).Alternatively,the energy in the afterglow is used (e.g.,Piran et al.2001).However,for GRB 011121,the energy in the SN component (scaling from the well-studied SN 1998bw)is likely to be comparable or even larger than that seen in the burst or the afterglow.In view of this,the apparent constancy of the γ-ray energy release is even more mysterious. 5 F l u x d e n s i t y (μJ y ) Day 13 ? 14 Day 23 ? 24 4000 5000 6000 7000 8000 9000 Wavelength (?) F l u x d e n s i t y (μJ y ) Day 27 ? 28 5000 6000 7000 8000 9000 Wavelength (?) x 10Day 76 ? 77 Fig.3.—The spectral ?ux distributions of the red bump at the time of the four HST epochs.The ?uxes are dereddened using A V =1.16mag.Spectral evolution,and more important,a turn-over in the spectra of the ?rst three epochs,are clearly seen.The peak of the turn-over (around 7200?A )corresponds to a peak in the red bump spectrum at ～5300?A .For comparison,we show a template broadband SN spectra (a dimmed version of SN 1998bw;solid curve)as it would appear at the redshift of GRB 011121and the associated 2σerrors (see text).The vertical error bars on the red bump re?ect the 1σstatistical uncertainty ?ux from only the red bump.There are large (～1mag)systematic uncertainties (e.g.,Galactic reddening,relative distance moduli between SN 1998bw and GRB 011121)in both the data and the model;these are suppressed for clarity. We thank S.Woosley,who,as referee,provided helpful insights toward the improvement of this work. A.Mac-Fadyen and E.Ramirez-Ruiz are acknowledged for their constructive comments on the paper.J.S.B.is a Fannie and John Hertz Foundation Fellow. F.A.H.acknowl-edges support from a Presidential Early Career award.S.R.K.and S.G.D.thank the NSF for support.R.S.is grateful for support from a NASA ATP grant.R.S.and T.J.G.acknowledge support from the Sherman Fairchild Foundation.J.C.W.acknowledges support from NASA grant NAG59302.KH is grateful for Ulysses support under JPL Contract 958056,and for IPN support under NASA Grants FDNAG 5-11451and NAG 5-17100.Support for Proposal number HST-GO-09180.01-A was provided by NASA through a grant from Space Telescope Science Insti-tute,which is operated by the Association of Universities for Research in Astronomy,Incorporated,under NASA Contract NAS5-26555. REFERENCES Berger,E.et al.2001,ApJ,556,556Bj¨o rnsson,G.,Hjorth,J.,Jakobsson,P.,Christensen,L.,and Holland,S.2001,ApJ (Letters ),552,L121 Bloom,J.S.,Kulkarni,S.R.,and Djorgovski,S.G.2002,AJ,123,1111 Bloom,J.S.et al.1999,Nature,401,453Carter,B.1992,ApJ (Letters ),391,L67 Castro-Tirado,A.J.et al.2001,A&A,370,398 Chevalier,R.A.and Li,Z.-Y.1999,ApJ (Letters ),520,L29Dar,A.and R′u jula,A.D.2002,astro-ph/0008474 Dermer,C.D.1996,in Gamma Ray Bursts:3rd Huntsville Symposium,ed.C.Kouveliotou,M.F.Briggs,&G.J.Fishman,Vol.384(Woodbury,New York:AIP),744–748Dolphin,A.E.2000,PASP,112,1397 Esin,A.A.and Blandford,R.2000,ApJ (Letters ),534,L151 Frail,D.A.et al.2001,ApJ (Letters ),562,L55—.2000,ApJ,534,559 Fruchter,A.,Vreeswijk,P.,Hook,R.,et al.2000,GCN notice 752Fryer,C.L.,Woosley,S.E.,and Hartmann,D.H.1999,ApJ,526,152 Fukugita,M.,Shimasaku,K.,and Ichikawa,T.1995,PASP,107,945 Galama,T.J.et al.2000,ApJ,536,185—.1998,Nature,395,670 Garnavich,P.M.,Holland,S.T.,Jha,S.,Kirshner,R.P.,Bersier,D.,and Stanek,K.Z.2002,GCN notice 1273H¨o ?ich,P.,Wheeler,J.C.,and Wang,L.1999,ApJ,521,179Hansen,B.M.S.1999,ApJ (Letters ),512,L117 6 Holtzman,J.A.,Burrows,C.J.,Casertano,S.,Hester,J.J., Trauger,J.T.,Watson,A.M.,and Worthey,G.1995,PASP, 107,1065 Infante,L.,Garnavich,P.M.,Stanek,K.Z.,and Wyrzykowski,L. 2001,GCN notice1152 Iwamoto,K.et al.1998,Nature,395,672 Khokhlov,A.M.,H¨o?ich,P.A.,Oran,E.S.,Wheeler,J.C.,Wang, L.,and Chtchelkanova,A.Y.1999,ApJ(Letters),524,L107 Kim,A.,Goobar,A.,and Perlmutter,S.1996,PASP,108,190 Lazzati,D.et al.2001,A&A,378,996 MacFadyen,A.I.and Woosley,S.E.1999,ApJ,524,262 MacFadyen,A.I.,Woosley,S.E.,and Heger,A.2001,ApJ,550, 410 Mazzali,P.A.et al.2002,ApJ(Letters),572,L61 McKenzie,E.H.and Schaefer,B.E.1999,PASP,111,964 Olsen,K.,Brown,M.,Schommer,R.,and Stubbs,C.2001,GCN Notice1157 Paczy′n ski,B.1988,ApJ,335,525 Perlmutter,S.et al.1997,ApJ,483,565 Piran,T.,Kumar,P.,Panaitescu,A.,and Piro,L.2001,ApJ (Letters),560,L167Piro,L.et al.2001,GCN notice1147 Price,P.A.,Schmidt,B.P.,and Kulkarni,S.R.2002a,GCN notice 1259 Price,P.A.et al.2002b,ApJ(Letters),572,L51(Paper II) Ramirez-Ruiz,E.,Dray,L.M.,Madau,P.,and Tout,C.A.2001, MNRAS,327,829 Reichart,D.E.1999,ApJ(Letters),521,L111 —.2001,ApJ,554,643 Sahu,K.C.et al.2000,ApJ,540,74 Schmidt,B.P.et al.1998,ApJ,507,46 Sokolov,V.V.2001,Bull.Spec.Astrophys.Obs.,51,38 Stanek,K.Z.,Garnavich,P.M.,and Wyrzykowski,L.2001,GCN notice1151 Stanek,K.Z.and Wyrzykowski,L.2001,GCN Notice1160 Stetson,P.B.1987,PASP,99,191 Vietri,M.and Stella,L.1998,ApJ(Letters),507,L45 Waxman,E.and Draine,B.T.2000,ApJ,537,796 Woosley,S.E.1993,ApJ,405,273 Wyrzykowski,L.,Stanek,K.Z.,and Garnavich,P.M.2001,GCN notice1150 7 Table1 Log of HST Imaging of the optical transient of GRB011121 Filter?t a Int.λe?b fν(λe?)b Vega Time Magnitude c (days)(sec)(?A)(μJy)(mag) a Mean time since GRB trigger on21.7828Nov2001UT. b In the fourth column,the e?ective wavelength of the?lter based upon the observed spectral?ux distribution of the transient at the given epoch.In the ?fth column,the?ux is given at this e?ective wavelength in an0′′.5radius.The observed count rate,corrected for CTE e?ects,was converted to?ux using the IRAF/SYNPHOT package.An input spectrum with fν=constant was?rst assumed.Then approximate spectral indices between each?lter were computed and then used to re-compute the?ux and the e?ective wavelength of the?l- ters.This bootstrapping converged after a few iterations.The HST photometry contains an unknown but small contribution from the host galaxy at the OT lo- cation.We attempted to estimate the contamination of the host at the transient position by measuring the host?ux in several apertures at approximate isopho- tal levels to the OT position.We estimate the contribution of the host galaxy to be fν(F450W)=(0.098±0.039)μJy,fν(F555W)=(0.087±0.027)μJy, fν(F702W)=(0.127±0.026)μJy,fν(F814W)=(0.209±0.059)μJy,and fν(F850LP)=(0.444±0.103)μJy.To correct these numbers to“in?nite aper- ture”multiply the?uxes by1.096(Holtzman et al.1995).These?uxes have not been corrected for Galactic or host extinction. c Tabulate d brightnesses in th e Vega magnitude system(B Vega=0.02mag, V Vega=0.03mag,R Vega=0.039mag,I Vega=0.035mag;Holtzman et al. 1995).Subtract0.1mag from these values to get the in?nite aperture brightness. These magnitudes have not been corrected for Galactic or host extinction.