1: \NeedsTeXFormat{LaTeX2e}[1996/06/01]
2:
3: \documentclass[cup6b,epsf]{cupbook}
4:
5: \usepackage{epsf}
6:
7: \def\note #1]{{\bf #1]}}
8:
9:
10: \def\note #1]{{\bf #1]}}
11: \def\dd{{\rm d}}
12: \def\be{{\bf e}}
13: \def\bu{{\bf u}}
14: \def\bx{{\bf x}}
15: \def\CK{{\cal K}}
16: \def\rt{r_{\rm t}}
17: \def\PP{{\cal P}}
18: \def\Eq #1{Eq.\ (\ref{#1})}
19:
20: \begin{document}
21:
22: \author{Thomas Matheson\\
23: Harvard-Smithsonian Center for Astrophysics \\
24: 60 Garden Street, Cambridge, MA 02138}
25:
26: \chapter{The First Direct Supernova/GRB Connection:
27: GRB~030329/SN~2003dh}
28:
29: {\it Observations of gamma-ray burst (GRB) afterglows have yielded
30: tantalizing hints that supernovae (SNe) and GRBs are related. The
31: case had been circumstantial, though, relying on irregularities in
32: the light curve or the colors of the afterglow. I will present
33: observations of the optical afterglow of GRB 030329. The early
34: spectra show a power-law continuum, consistent with other GRB
35: afterglows. After approximately one week, broad peaks in the
36: spectrum developed that were remarkably similar to those seen in
37: the spectra of the peculiar Type Ic SN 1998bw. This is the first
38: direct, spectroscopic confirmation that at least some GRBs arise
39: from SNe.
40: }
41:
42: \section{Introduction}
43:
44: The mechanism that produces gamma-ray bursts (GRBs) has been the
45: subject of considerable speculation during the four decades since
46: their discovery (see M{\'e}sz{\'a}ros 2002 for a recent review of the
47: theories of GRBs). Optical afterglows (e.g.,~GRB~970228: Groot et
48: al.~1997; van Paradijs et al.~1997) opened a new window on the field
49: (see, e.g., van Paradijs, Kouveliotou, \& Wijers 2000). Subsequent
50: studies of other bursts yielded the redshifts of several GRBs
51: (e.g.,~GRB~970508: Metzger et al.~1997), providing definitive evidence
52: for their cosmological origin.
53:
54: Models that invoked supernovae (SNe) to explain GRBs were proposed
55: from the very beginning (e.g., Colgate 1968; Woosley 1993; Woosley \&
56: MacFadyen 1999). A strong hint was provided by GRB~980425. In this
57: case, no traditional GRB optical afterglow was seen, but a supernova,
58: SN~1998bw, was found in the error box of the GRB (Galama et
59: al. 1998a). The SN was classified as a Type Ic (Patat \& Piemonte
60: 1998), but it was unusual, with high expansion velocities (Patat et
61: al. 2001). Other SNe with high expansion velocities (and usually
62: large luminosity as well) such as SN~1997ef and SN~2002ap are
63: sometimes referred to as ``hypernovae'' (see, e.g., Iwamoto et
64: al. 1998, 2000).
65:
66: The redshift of a typical GRB is $z \approx 1$, implying that a
67: supernova component underlying an optical afterglow would be
68: difficult to detect. At $z \approx 1$, even a bright core-collapse
69: event would peak at $R > 23$ mag. Nevertheless, late-time deviations
70: from the power-law decline typically observed for optical afterglows
71: have been seen and these bumps in the light curves have been
72: interpreted as evidence for supernovae (for a recent summary, see
73: Bloom 2003). Perhaps the best evidence that classical, long-duration
74: gamma-ray bursts are generated by core-collapse supernovae was
75: provided by GRB~011121. It was at $z = 0.36$, so the supernova
76: component would have been relatively bright. A bump in the light
77: curve was observed both from the ground and with \emph{HST} (Garnavich
78: et al.~2003a; Bloom et al.~2002). The color changes in the light
79: curve of GRB~011121 were also consistent with a supernova (designated
80: SN~2001ke), but a spectrum obtained by Garnavich et al. (2003a) during
81: the time that the bump was apparent did not show any features that
82: could be definitively identified as originating from a supernova. The
83: detection of a clear spectroscopic supernova signature was for the
84: first time reported for the GRB~030329 by Matheson et al. (2003a,
85: 2003b), Garnavich et al. (2003b, 2003c), Chornock et al. (2003), and
86: Stanek et al.~(2003a). Hjorth et al. (2003) also presented
87: spectroscopic data obtained with the VLT. In addition, Kawabata et
88: al. (2003) obtained a spectrum of SN~2003dh with the Subaru
89: telescope.
90:
91: The extremely bright GRB~030329 was detected by instruments aboard
92: \emph{HETE II} at 11:37:14.67 (UT is used throughout this paper) on
93: 2003 March~29 (Vanderspek et al.~2003). Due to the brightness of the
94: afterglow, observations of the optical transient (OT) were extensive,
95: making it most likely the best-observed afterglow so far. From the
96: moment the low redshift of 0.1685 for the GRB~030329 was announced
97: (Greiner et al.~2003), we started organizing a campaign of
98: spectroscopic and photometric follow-up of the afterglow and later the
99: possible associated supernova. Stanek et al.~(2003a) reported the
100: first results of this campaign, namely a clear spectroscopic detection
101: of a SN~1998bw-like supernova in the early spectra, designated
102: SN~2003dh (Garnavich et al. 2003c). In this paper, I describe the
103: evidence for the supernova in the spectroscopy during the first two
104: months. For a more complete discussion, see Matheson et al.~(2003c).
105:
106: \begin{figure}
107: \begin{center}
108: \leavevmode\epsfxsize=10cm \epsfbox{fig1.eps}
109: \end{center}
110: %Figure caption
111: \caption{Evolution of the GRB~030329/SN~2003dh spectrum, from
112: March~30.23 UT (0.75 days after the burst), to April 10.14 UT (11.66
113: days after the burst). The early spectra consist of a power-law
114: continuum with narrow emission lines originating from {\sc{H II}}
115: regions in the host galaxy at $z = 0.1685$. Spectra taken after
116: $\Delta T=6.66$ days show the development of broad peaks characteristic of
117: a supernova.
118: }
119: \label{fig:grb-all}
120: \end{figure}
121:
122:
123:
124: \section{Spectra}
125:
126: The brightness of the OT allowed us to observe the OT each of the 12
127: nights between March 30 and April 10 UT, mostly with the MMT 6.5-m,
128: but also with the Magellan 6.5-m, Lick Observatory 3-m, LCO du~Pont
129: 2.5-m, and FLWO 1.5-m telescopes. This provided a unique opportunity
130: to look for spectroscopic evolution over many nights. The early
131: spectra of the OT of GRB~030329 (top of Figure \ref{fig:grb-all}) consist
132: of a power-law continuum typical of GRB afterglows, with narrow
133: emission features identifiable as H$\alpha$, [{\sc{O III}}]
134: $\lambda\lambda$4959, 5007, H$\beta$, and [{\sc{O II}}] $\lambda$3727
135: at $z = 0.1685$ (Greiner et al.~2003; Caldwell et al.~2003) probably
136: from {\sc{H II}} regions in the host galaxy.
137:
138: Beginning at $\Delta T=7.67$ days, our spectra deviated from the pure
139: power-law continuum. Broad peaks in flux, characteristic of a
140: supernova, appeared. The broad bumps are seen at approximately
141: 5000 \AA\ and 4200 \AA\ (rest frame). At that time, the spectrum of
142: GRB~030329 looked similar to that of the peculiar Type Ic SN~1998bw
143: a week before maximum light (Patat et al. 2001) superposed on a
144: typical afterglow continuum. Over the next few days the SN features
145: became more prominent as the afterglow faded and the SN brightened
146: toward maximum.
147:
148:
149: \begin{figure}
150: \begin{center}
151: \leavevmode\epsfxsize=10cm \epsfbox{fig2.eps}
152: \end{center}
153: %Figure caption
154: \caption{Evolution of the GRB~030329/SN~2003dh spectrum, from April
155: 24.28 UT (25.8 days after the burst), to May 24.38 (55.9 days after
156: the burst). The power-law contribution decreases and the spectra
157: become more red as the SN component begins to dominate, although the
158: upturn at blue wavelengths may still be the power law. The broad
159: features of a supernova are readily apparent, and the overall spectrum
160: continues to resemble that of SN~1998bw several days after maximum.
161: }
162: \label{fig:spectra-later}
163: \end{figure}
164:
165: Later spectra obtained on April 24.28, May 2.05, May 4.01, and May
166: 24.38 continue to show the characteristics of a supernova. As
167: the power-law continuum of the GRB afterglow fades, the supernova
168: spectrum rises, becoming the dominant component of the overall
169: spectrum (Figure \ref{fig:spectra-later}).
170:
171:
172: \section{Separating the GRB from the Supernova}
173:
174: To explore the nature of the supernova underlying the OT, we modeled
175: the spectrum as the sum of a power-law continuum and a peculiar Type
176: Ic SN. Specifically, we chose for comparison SN~1998bw (Patat et
177: al. 2001), SN~1997ef (Iwamoto et al. 2000), and SN~2002ap (using our
178: own as yet unpublished spectra, but see, e.g., Kinugasa et al. 2002;
179: Foley et al. 2003). We had 62 spectra of these three SNe, spanning
180: the epochs of seven days before maximum to several weeks past. For
181: the power-law continuum, we chose to use one of our early spectra to
182: represent the afterglow of the GRB. The spectrum at time $\Delta
183: T=5.80$ days was of high signal-to-noise ratio (S/N), and suffers from
184: little fringing at the red end. Therefore, we smoothed this spectrum
185: to provide the fiducial power-law continuum of the OT for our model.
186:
187: To find the best match with a supernova spectrum, we compared each
188: spectrum of the afterglow with the sum of the fiducial continuum and a
189: spectrum of one of the SNe in the sample. We performed a
190: least-squares fit, allowing the fraction of continuum and SN to vary,
191: finding the best combination of continuum and SN for each of the SN
192: spectra. The minimum least-squares deviation within this set was then
193: taken as the best SN match for that epoch of OT observation. Figure
194: \ref{fig:snfrac} shows the relative contribution to the OT spectrum by
195: the underlying SN in the $B$ and $R$ bands as a function of $\Delta
196: T$.
197:
198: \begin{figure}
199:
200: \begin{center}
201: \leavevmode\epsfxsize=10cm \epsfbox{fig3.eps}
202: \end{center}
203: %Figure caption
204: \caption{Relative contribution of a supernova spectrum to the
205: GRB~030329/SN~2003dh afterglow as a function of time in the $B$
206: (\emph{open circles}) and $R$ (\emph{filled squares}) bands. Using
207: the technique described in the text, we derive a best fit to the
208: afterglow spectrum at each epoch with the fiducial power-law continuum
209: and the closest match from our set of peculiar SNe Ic. We then
210: synthesize the relative $B$-band and $R$-band contributions. There is
211: some scatter for the early epochs due to noise in the spectra, but a
212: clear deviation is evident starting at $\Delta T=7.67$ days, with a
213: subsequent rapid increase in the fraction of the overall spectrum
214: contributed by the SN. Errors are estimated from the scatter when the
215: SN component is close to zero ($\Delta T < 6$ days) and from the scale
216: of the error in the least-squares minimization.
217: }
218:
219: \label{fig:snfrac}
220: \end{figure}
221:
222: Within the uncertainties of our fit, the SN fraction is consistent
223: with zero for the first few days after the GRB. At $\Delta T=7.67$
224: days, the SN begins to appear in the spectrum, without strong evidence
225: for a supernova component before this. When the fit indicates the
226: presence of a supernova, the best match is almost always SN~1998bw.
227: The only exceptions to this are from nights when the spectrum of the
228: OT are extremely noisy, implying that less weight should be given to
229: those results. The least-squares deviation for the spectra that do
230: not match SN~1998bw is also much larger.
231:
232: Our best spectrum (i.e., with the highest S/N) from this time when the
233: SN features begin to appear is at $\Delta T=9.67$ days. For that
234: epoch, our best fit is 74\% continuum and 26\% SN~1998bw (at day $-6$
235: relative to SN $B$-band maximum). The next-best fit is SN~1998bw at
236: day $-$7. Using a different early epoch to define the reference
237: continuum does not alter these results significantly. It causes
238: slight changes in the relative percentages, but the same SN spectrum
239: still produces the best fit, albeit with a larger least-squares
240: deviation.
241:
242: The SN fraction contributing to the total spectrum increases steadily
243: with time. By $\Delta T=25.8$ days, the SN fraction is $\sim$ 61\%,
244: with the best-fit SN being SN~1998bw at day +6 (Figure \ref{fig:day25}).
245: The SN percentage at $\Delta T=33.6$ days is still about 63\%, but the
246: best match is now SN~1998bw at day +13. The rest-frame time
247: difference between $\Delta T=9.67$ days and $\Delta T=25.8$ days is
248: 13.8 days ($z = 0.1685$). For the best-fit SN spectra from those
249: epochs, SN~1998bw at day $-$6 and SN~1998bw at day +6 respectively,
250: the time difference is 12 days. The rest-frame time difference
251: between $\Delta T=25.8$ days and $\Delta T=33.6$ days is 6.7 days,
252: with a time difference between the best-fit spectra for those epochs
253: of 7 days. The spectral evolution determined from these fits
254: indicates that SN~2003dh follows SN~1998bw closely, and it is not as
255: similar to SN~1997ef or SN~2002ap. The analysis by Kawabata et
256: al. (2003) of their May 10 spectrum gives a phase for the spectrum of
257: SN~2003dh that is consistent with our dates, although they do consider
258: SN~1997ef as a viable alternative to SN~1998bw as a match for the SN
259: component in the afterglow.
260:
261: \begin{figure}
262: \begin{center}
263: \leavevmode\epsfxsize=10cm \epsfbox{fig4.eps}
264: \end{center}
265: %Figure caption
266: \caption{Observed spectrum (\emph{thin line}) of the
267: GRB~030329/SN~2003dh afterglow at $\Delta T=25.8$ days. The model spectrum
268: (\emph{thick line}) consists of 39\% continuum and 61\% SN~1998bw from
269: 6 days after maximum.
270: }
271: \label{fig:day25}
272: \end{figure}
273:
274:
275: The spectra of SN~1998bw (and other highly energetic SNe) are not
276: simple to interpret. The high expansion velocities result in many
277: overlapping lines so that identification of specific line features is
278: problematic for the early phases of spectral evolution (see, e.g.,
279: Iwamoto et al. 1998; Stathakis et al. 2000; Nakamura et al. 2001;
280: Patat et al. 2001). This includes spectra up to two weeks after
281: maximum, approximately the same epochs covered by our spectra of
282: SN~2003dh. In fact, as Iwamoto et al. (1998) showed, the spectra at
283: these phases do not show line features. The peaks in the spectra are
284: due to gaps in opacity, not individual spectral lines. Detailed
285: modeling of the spectra can reveal some aspects of the composition of
286: the ejecta (Mazzali et al. 2003).
287:
288: If the $\Delta T=9.67$ days spectrum for the afterglow does match
289: SN~1998bw at day $-$6, then limits can be placed on the timing of the
290: supernova explosion relative to the GRB. The rest-frame time for
291: $\Delta T=9.67$ days is 8.2 days, implying that the time of the GRB
292: would correspond to $\sim$14 days before maximum for the SN. The rise
293: times of SNe Ic are not well determined, especially for the small subset
294: of peculiar ones. Stritzinger et al. (2002) found the rise time of
295: the Type Ib/c SN~1999ex was $\sim$18 days (in the $B$ band), while
296: Richmond et al. (1996) reported a rise time of $\sim$12 days (in the
297: $V$ band) for the Type Ic SN~1994I. A rise time of $\sim$14 days for
298: SN~2003dh is certainly a reasonable number. It also makes it
299: extremely unlikely that the SN exploded significantly earlier or later
300: than the time of the GRB, most likely within $\pm 2$ days of the GRB
301: itself.
302:
303: \begin{figure}
304: \begin{center}
305: \leavevmode\epsfxsize=10cm \epsfbox{fig5.eps}
306: \end{center}
307: %Figure caption
308: \caption{Decomposition of the OT $R$-band light curve into the
309: supernova (\emph{dotted line}) and the power-law continuum
310: (\emph{dashed line}). As the light curve model for the supernova, we
311: took the $V$-band light curve of SN~1998bw (Galama et al. 1998a, b)
312: stretched by $(1+z)=1.1685$ and shifted in magnitude. The resulting
313: supernova light curve peaks at an apparent magnitude of $m_R=20.4$. No
314: offset in time has been applied between the GRB and the supernova. To
315: constrain the continuum, information from the spectral decomposition
316: was used (\emph{big open circles}).
317: }
318: \label{fig:decomp}
319: \end{figure}
320:
321:
322: The totality of data contained in this paper allows us to attempt to
323: decompose the light curve of the OT into the supernova and the
324: afterglow (power-law) component. From the spectral decomposition
325: procedure described above, we have the fraction of light in the
326: $BR$-bands for both components at various times, assuming that the
327: spectrum of the afterglow did not evolve since $\Delta T=5.64$ days.
328: As we find that the spectral evolution is remarkably close to that of
329: SN~1998bw, we model the $R$-band supernova component with the $V$-band
330: light curve of SN~1998bw (Galama et al. 1998a, b) stretched by
331: $(1+z)=1.1685$ and shifted in magnitude to obtain a good fit. The
332: afterglow component is fit by using the early points starting at
333: $\Delta T=5.64$ days with late points obtained via the spectral
334: decomposition. This can be done in both in the $B$ and in the $R$-band
335: and leads to consistent results, indicating that our assumption of the
336: afterglow not evolving in color at later times is indeed valid.
337:
338:
339: The result of the decomposition of the OT $R$-band light curve into
340: the supernova and the power-law continuum is shown in Figure
341: \ref{fig:decomp}. The overall fit is remarkably good, given the
342: assumptions (such as using the stretched $V$-band light curve of
343: SN~1998bw as a proxy for the~SN 2003dh $R$-band light curve). No time
344: offset between the supernova and the GRB was applied, and given how
345: good the fit is, we decided not to explore time offset as an
346: additional parameter. Introducing such an additional parameter would
347: most likely result in a somewhat better fit (indeed, we find that to
348: be the case for $\delta t\approx -2$ days), but this could easily be
349: an artifact with no physical significance, purely due to small
350: differences between SN~1998bw and SN~2003dh. At this point the
351: assumption that the GRB and the SN happened at the same time seems
352: most natural.
353:
354:
355:
356: \section{Summary}
357:
358: The spectroscopy of the optical afterglow of GRB~030329, as first
359: shown by Stanek et al. (2003a), provided direct evidence that at least
360: some of the long-burst GRBs are related to core-collapse SNe. We have
361: shown with a larger set of data that the SN component is similar
362: to SN~1998bw, an unusual Type Ic SN. It is not clear yet whether all
363: long-burst GRBs arise from SNe. Catching another GRB at a redshift
364: this low is unlikely, but large telescopes may be able to discern SNe
365: in some of the relatively nearby bursts. With this one example,
366: though, we now have solid evidence that some GRBs and SNe have the
367: same progenitors.
368:
369: \bigskip\noindent
370:
371: {\it Acknowledgments} Kris Stanek and Peter
372: Garnavich were very supportive and equal colleagues in the research
373: described here. I would like to thank the many observers who
374: sacrificed their time to observe this GRB (see Matheson et al. 2003c
375: for a full list).
376:
377: \begin{thereferences}{}
378:
379: \makeatletter
380: \renewcommand{\@biblabel}[1]{\hfill}
381:
382: \bibitem[Bloom (2003)]{bloom03} Bloom, J.~S., 2003, in \emph{Gamma-Ray Bursts
383: in the Afterglow Era}, ed. M. Feroci et al. (San Francisco: ASP), 1.
384:
385: \bibitem[Bloom et al.(2002)]{bloom02b} Bloom, J.~S., et al., 2002, \emph{Astrophys. J.},
386: {\bf 572}, L45.
387:
388: \bibitem[Caldwell et al.(2003)]{caldwell03} Caldwell, N., Garnavich,
389: P., Holland, S., Matheson, T., \& Stanek, K.Z., 2003, \emph{GCN Circ.}~2053.
390:
391: \bibitem[Chornock et al. (2003)]{chornock03} Chornock, R., Foley,
392: R.~J., Filippenko, A.~V., Papenkova, M., \& Weisz, D., 2003, \emph{GCN
393: Circ.}~2131.
394:
395: \bibitem[Colgate(1968)]{colgate68} Colgate, S.~A., 1968, \emph{Canadian
396: J. Phys.}, {\bf 46}, 476.
397:
398:
399: \bibitem[Foley et al.(2003)]{foley03} Foley, R.~J., et al., 2003,
400: \emph{Pub. Astron. Soc. Pac.}, in press (astro-ph/0307136).
401:
402: \bibitem[Galama et al.(1998a)]{galama98a} Galama, T.~J.,~et al.,
403: 1998a, \emph{Nature}, {\bf 395}, 670.
404:
405: \bibitem[Galama et al.(1998b)]{Galama98b} Galama, T.~J., et al., 1998b,
406: \emph{Astrophys. J.}, {\bf 497}, L13.
407:
408: \bibitem[Garnavich et al.(2003a)]{garnavich03a} Garnavich, P.~M., et
409: al., 2003a, \emph{Astrophys. J.}, {\bf 582}, 924.
410:
411: \bibitem[Garnavich et al.(2003b)]{garnavich03b} Garnavich, P., et al..
412: 2003b, \emph{IAU Circ.} 8108.
413:
414: \bibitem[Garnavich et al.(2003c)]{garnavich03c} Garnavich, P.,
415: Matheson, T., Olszewski, E.~W., Harding, P., \& Stanek, K.~Z., 2003c,
416: \emph{IAU Circ.} 8114.
417:
418: \bibitem[Greiner et al. (2003)]{greiner03} Greiner, J., et al., 2003,
419: \emph{GCN Circ.}~2020.
420:
421: \bibitem[Groot et al.(1997)]{groot97} Groot, P.~J., et al., 1997,
422: \emph{IAU Circ.} 6584.
423:
424: \bibitem[Hjorth et al.(2003)]{hjorth03} Hjorth, J., et al., 2003, \emph{Nature},
425: {\bf 423}, 847.
426:
427: \bibitem[Iwamoto et al.(1998)]{iwamoto98} Iwamoto, K.,~et al.,
428: 1998, \emph{Nature}, {\bf 395}, 672.
429:
430: \bibitem[Iwamoto et al.(2000)]{iwamoto00} Iwamoto, K., et al., 2000,
431: \emph{Astrophys. J.}, {\bf 534}, 660.
432:
433: \bibitem[Kawabata et al.(2003)]{kawabata03} Kawabata, K.~S., et
434: al., 2003, \emph{Astrophys. J.}l, {\bf 593}, L19.
435:
436: \bibitem[Kinugasa et al.(2002)]{kinugasa02} Kinugasa, K., et al., 2002,
437: \emph{Astrophys. J.}, {\bf 577}, L97.
438:
439: \bibitem[Matheson et al.(2003a)]{matheson03a} Matheson, T., et al.,
440: 2003a, \emph{GCN Circ.}~2107.
441:
442: \bibitem[Matheson et al.(2003b)]{matheson03b} Matheson, T., et
443: al., 2003b, \emph{GCN Circ.}~2120.
444:
445: \bibitem[Matheson et al.(2003c)]{matheson03c} Matheson, T., et
446: al., 2003c, \emph{Astrophys. J.}, in press (astro-ph/0307435).
447:
448: \bibitem[Mazzali et al.(2003)]{mazzali03} Mazzali, P.~A., et
449: al., 2003, \emph{Astrophys. J.}, submitted (astro-ph/0309555).
450:
451: \bibitem[M{\'e}sz{\'a}ros(2002)]{M2002} M{\'e}sz{\'a}ros, P., 2002,
452: \emph{Ann. Rev. Astron. Astrophys.}, {\bf 40}, 137.
453:
454: \bibitem[Metzger et al.(1997)]{metzger97} Metzger, M.~R., et al., 1997, \emph{Nature}, {\bf 387}, 878.
455:
456: \bibitem[Nakamura et al.(2001)]{nakamura01} Nakamura, T., Mazzali,
457: P.~A., Nomoto, K., \& Iwamoto, K., 2001, \emph{Astrophys. J.}, {\bf 550}, 991.
458:
459: \bibitem[Patat \& Piemonte (1998)]{patat98} Patat, F., \& Piemonte
460: A., 1998, \emph{IAU Circ.} 6918.
461:
462: \bibitem[Patat et al.(2001)]{patat01} Patat, F., et al., 2001, \emph{Astrophys. J.},
463: {\bf 555}, 900.
464:
465: \bibitem[Richmond et al.(1996)]{richmond96} Richmond, M.~W.,~et
466: al., 1996, \emph{Astron. J.}, {\bf 111}, 327.
467:
468: \bibitem[Stanek et al.(2003a)]{stanek03a} Stanek, K.~Z., et al., 2003a,
469: \emph{Astrophys. J.}, {\bf 591}, L17.
470:
471: \bibitem[Stathakis et al.(2000)]{stathakis00} Stathakis, R.~A., et
472: al., 2000, \emph{Mon. Not. R. Astr. Soc.}, {\bf 314}, 807.
473:
474: \bibitem[Stritzinger et al.(2002)]{stritzinger02} Stritzinger, M.,~et
475: al., 2002, \emph{Astron. J.}, {\bf 124}, 2100.
476:
477: \bibitem[Vanderspek et al.(2003)]{vanderspek03} Vanderspek, R., et
478: al., 2003, \emph{GCN Circ.}~1997.
479:
480: \bibitem[van Paradijs, Kouveliotou, \& Wijers(2000)]{paradijs00}
481: van Paradijs, J., Kouveliotou, C., \& Wijers, R.~A.~M.~J., 2000, \emph{Ann. Rev. Astron. Astrophys.}, {\bf 38},
482: 379.
483:
484: \bibitem[van Paradijs et al. (1997)]{vanparadijs97} van Paradijs, J.,
485: et al., 1997, \emph{Nature}, {\bf 386}, 686.
486:
487: \bibitem[Woosley (1993)]{woosley93} Woosley, S.~E., 1993, \emph{Astrophys. J.}, {\bf 405}, 273.
488:
489: \bibitem[Woosley \& MacFadyen(1999)]{woosley99a} Woosley, S.~E.,~\&
490: MacFadyen, A.~I., 1999, \emph{Astron. Astrophys.}, {\bf 138}, 499.
491:
492: \end{thereferences}
493:
494: \end{document}
495: