1: \documentclass{emulateapj}
2: %\documentclass[12pt,preprint]{aastex}
3: \slugcomment{Accepted for publication in ApJ}
4:
5: \begin{document}
6: \title{Optical afterglows of short Gamma-ray Bursts and GRB 040924}
7: %\footnote{}
8:
9: \author{Y. Z. Fan$^{1,2,3}$, Bing Zhang$^{1}$, Shiho Kobayashi$^{4,5}$ and
10: Peter M\'{e}sz\'{a}ros$^{4,5}$}
11: \affil{$^1$ Dept. of Physics, University of Nevada, Las Vegas, NV
12: 89154, USA.\\
13: $^2$ Purple Mountain Observatory, Chinese Academy of
14: Science, Nanjing 210008, China.\\
15: $^3$ National Astronomical Observatories, Chinese Academy of
16: Sciences, Beijing, 100012, China.\\
17: $^4$ Dept. of Astronomy and Astrophysics, Pennsylvania State
18: University, 525 Davey Laboratory, University Park, PA 16802, USA.\\
19: $^5$ Dept. of Physics, Pennsylvania State
20: University, 104 Davey Laboratory, University Park, PA 16802, USA.}
21:
22: \begin{abstract}
23: Short-duration Gamma-ray bursts (GRBs) ($\leq 2{\rm s}$) have remained
24: a mystery due to the lack of afterglow detection until
25: recently. The models to interpret short GRBs invoke distinct
26: progenitor scenarios. Here we present a generic analysis of short GRB
27: afterglows, and calculate the optical lightcurves of short GRBs
28: within the framework of different progenitor models. We show that all
29: these optical afterglows are bright enough to be detected by the
30: Ultraviolet and Optical Telescope (UVOT) on board the {\em Swift}
31: observatory, and that different models could be distinguished with a
32: well-monitored lightcurve. We also model the afterglow data of the
33: recently discovered short burst GRB 040924. We find that the
34: limited data are consistent with a low medium density environment
35: which is consistent with the pre-concept of the compact-star merger
36: progenitor model, although the models with a
37: collapsar progenitor are not ruled out.
38: \end{abstract}
39:
40: \keywords{Gamma Rays: bursts$-$ISM: jets and outflows--radiation
41: mechanisms: nonthermal}
42:
43: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
44: \section{Introduction}
45: In the past several years great advances have been been made in
46: revealing the nature of Gamma-ray Bursts (GRBs) of relatively long
47: duration, i.e. $T_{90}>2{\rm s}$ (e.g. M\'{e}sz\'{a}ros 2002; Zhang \&
48: M\'{e}sz\'{a}ros 2004 for recent reviews). However, another category
49: of GRBs, i.e. those with short durations (i.e. $T_{90}<2{\rm s}$),
50: which comprise about 1/3 of the total GRB population, have remained
51: as mysterious as long GRBs were before 1997. This has been mainly
52: due to the lack of afterglow detections for short GRBs, until very
53: recently.
54:
55: The leading progenitor model for short GRBs invokes merger of two
56: compact objects (e.g. NS-NS merger or BH-NS merger, Eichler et al. 1989;
57: Paczy\'{n}ski 1991; Narayan et al. 1992; M\'{e}sz\'{a}ros \& Rees
58: 1992), which has been found suitable to interpret many short GRB
59: properties (Ruffert et al. 1997; Popham et al. 1999; Perna
60: \& Belczynski 2002; Rosswog et al. 2003; Aloy et al. 2004).
61: In this scenario, the burst
62: site is expected to have a large offset from the host galaxy due to
63: asymetric kicks during the birth of NSs (Bloom et al.
64: 1999; but see Belczynski et al. 2002), so that the number
65: density of the external medium in the GRB environment is low,
66: typically $\sim 10^{-2}{\rm cm^{-3}}$. Alternatively, with the
67: increasing evidence that long GRB progenitors are collapsars,
68: it has been suggested that short GRBs may also be associated with
69: collapsars, with either a less energetic jet (i.e. short emerging
70: model, Zhang, Woosley \& MacFadyen 2003) or a jet composed of many
71: subjets seen by an off-axis observer looking into one or a few
72: subjet(s) (subjets model, Yamazaki, Ioka \& Nakamura 2004). If this
73: is the case, the environment around the progenitor should be similar
74: to that of long GRBs, which is either a constant density medium
75: (e.g. Panaitescu \& Kumar 2002; Yost et al. 2003) with
76: ISM number density $n\sim 1{\rm cm^{-3}}$ or a prestellar wind
77: (e.g. Chevalier \& Li 2000). Other possibilities for the origin of
78: short GRBs have been proposed within the cylindrical jet model
79: (Wang et al. 2004) and the Poynting-flux dominated GRB model
80: (Lyutikov \& Blandford 2003).
81:
82: Within the standard afterglow model and adopting a typical compact star
83: merger environment, the forward shock afterglow emission of short GRBs
84: have been calculated by Panaitescu et al. (2001), Perna \& Belczynski
85: 2002, and Li et al. (2003). Panaitescu et al. (2001) have shown
86: that the afterglows of short GRBs are faint, and they are likely to be
87: most easily detected in the X-ray band. Li et al. (2003) considered
88: possible $e^\pm$ pair loading and evaluated its possible observational
89: signature. In this work, we present a generic treatment of short GRB
90: optical afterglows which differs from the previous ones by including
91: both the forward and the reverse shock emission, a crucial ingredient
92: for characterizing the early afterglow light curve and the
93: spectrum. The model is applied to various progenitor models and sample
94: lightcurves are calculated which are compared against Swift UVOT
95: sensitivity (\S2). Lately, a short, soft burst GRB 040924 was located
96: by HETE-2, which led to the discovery of its optical afterglow (Fox \&
97: Moon 2004). We also apply the model to fit the afterglow data of this
98: burst (\S3).
99:
100: \section{The afterglow of short $\gamma-$ray bursts}
101:
102: In the standard afterglow model for a fireball interacting with a
103: constant density medium (e.g., Sari, Piran \& Narayan
104: 1998), for the forward shock (FS) emission, the cooling frequency
105: $\nu_{\rm c}^{\rm f}$, the typical synchrotron frequency $\nu_{\rm
106: m}^{\rm f}$ and the maximum spectral flux $F_{\rm \nu,max}^{\rm f}$
107: read
108: \begin{equation}
109: \nu_{\rm c}^{\rm f}=4.3\times 10^{17}{\rm
110: Hz}~E_{51}^{-1/2}\epsilon_{\rm B,-2}^{-3/2}n_{-2}^{-1}{t}_{\rm
111: d}^{-1/2}({2\over 1+z}),
112: \end{equation}
113: \begin{equation}
114: \nu_{\rm m}^{\rm f}=3.9\times 10^{11}{\rm
115: Hz}~E_{51}^{1\over 2}\epsilon_{\rm B,-2}^{1\over 2}\epsilon_{\rm
116: e,-0.5}^2{t}_{\rm d}^{-{3\over 2}}[{13(p-2)\over 3(p-1)}]^2({2\over 1+z}),
117: \end{equation}
118: \begin{equation}
119: F_{\rm \nu,max}^{\rm f}=8.3 {\rm \mu Jy}~E_{51}\epsilon_{\rm
120: B,-2}^{1/2}n_{-2}^{1/2}D_{28.34}^{-2}({1+z\over 2}),
121: \end{equation}
122: where $E$ is the isotropic energy of the outflow, $\epsilon_{\rm e}$ and
123: $\epsilon_{\rm B}$ are the fractions of the shock energy given to the
124: magnetic field and electron at the shock, respectively, $n$ is the
125: number density of the external medium, $p\sim2.3$ is the power-law
126: distribution index of shocked electrons, $D$ is the luminosity
127: distance, and $z$ is the redshift. Hereafter $t=t_{\rm obs}/(1+z)$ denotes
128: the observer's time corrected for the cosmological time dilation
129: effect, and $t_{\rm d}$ is in unit of day.
130: The superscript ``f'' (``r'') represent the forward
131: (reverse) shock emission respectively. Throughout this work, we adpot
132: the convention $Q_{\rm x}=Q/10^{\rm x}$ using cgs units.
133: We have normalized the parameters to typical values of short GRBs.
134: The above equations apply to an isotropic fireball, or to a jet
135: with opening angle $\theta_0$ when the bulk Lorentz factor $\gamma >
136: 1/(\sqrt{3}\theta_0)$, so that $\gamma\approx
137: 8.2E_{51}^{1/8}n_{-2}^{-1/8}{t}_{\rm d}^{-3/8}$ is satisfied.
138: If sideways expansion is important, for $\gamma\leq
139: 1/(\sqrt{3}\theta_0)$, one has $\gamma=(\sqrt{3}\theta_0)^{-1} (t_{\rm
140: d}/t_{\rm 0,d})^{-1/2}$, $F_{\rm \nu,max(J_{\rm s})}^{\rm f}=F_{\rm
141: \nu,max}^{\rm f}(t_{\rm d}/t_{\rm
142: 0,d})^{\rm -1}$, $\nu_{\rm c(J_{\rm s})}^{\rm f}\approx \nu_{\rm
143: c}^{\rm f}(t_{\rm 0,d})$ and $\nu_{\rm m(J_{\rm s})}^{\rm f}\approx
144: \nu_{\rm m}^{\rm f}(t_{\rm 0,d})({t}_{\rm d}/t_{\rm 0,d})^{-2}$
145: (Rhoads 1999; Sari, Piran \& Halpern 1999). If sideways expansion is
146: unimportant,
147: equations (1-2) still hold and equation (3) should be replaced by
148: $F_{\rm \nu,max(J)}^{\rm f}\approx F_{\rm \nu,max}^{\rm f}(t_{\rm
149: d}/t_{\rm 0,d})^{-3/4}$. Here the subsecript $J$ ($J_{\rm s}$)
150: represnts a jet without (with) significant sideways expansion,
151: respectively. During the reverse shock crossing process, the bulk LF
152: of the ejecta is nearly a constant if the reverse shock is
153: non-relativistic (which is the case for short bursts). We have $F_{\rm
154: \nu, max}^{\rm f}\propto t^3$, $\nu_{\rm c}^{\rm f}\propto t^{-2}$ and
155: $\nu^{\rm f}_{\rm m}$ is independent on $t$.
156:
157: The time when RS crosses the shell can be
158: estimated by $t_{\rm \times}=\max[t_{\rm dec}, T_{90,\rm obs}/(1+z)]$
159: (Kobayashi, Piran \& Sari 1999). The typical
160: duration of short bursts is $T_{90, \rm obs}\sim 0.2{\rm s}$, which is much
161: smaller than the deceleration time $t_{\rm dec}$ for the ISM case.
162: We therefore have a typical thin-shell regime.
163: the RS is only mildly-relativistic at the shock crossing time (e.g.
164: Sari \& Piran 1999; Kobayashi 2000). The typical deceleration radius
165: is defined as $R_{\rm dec}\approx 5.6\times 10^{16}{\rm cm}~E_{51}^{1/3}n_{\rm
166: -2}^{-1/3}\eta_{2.5}^{-2/3}$ (Rees \& M\'{e}sz\'{a}ros 1992),
167: where $\eta\sim 300$ is the initial Lorentz factor (LF) of the outflow.
168: At $R_{\rm dec}$, the LF of the outflow drops to
169: $\gamma_{\times}=\gamma_{\rm dec}\sim 0.6 \eta$, so that
170: ${t}_{\rm dec}\approx R_{\rm dec}/2\gamma_{\rm dec}^2c=30{\rm s}~
171: E_{51}^{1/3} n_{-2}^{-1/3} \eta_{2.5}^{-8/3}$.
172:
173: At ${t}_{\rm \times}=t_{\rm dec}$, the LF of the decelerated outflow
174: relative to the initial one is
175: $\gamma_{34,\times}\approx (\eta/\gamma_{\rm \times}+\gamma_{\rm
176: \times}/\eta)/2=1.13$.
177: The typical frequency of the RS emission can be estimated by
178: \begin{equation}
179: \nu_{\rm m}^{\rm r}(t_{\rm \times})={\cal R}_{\rm
180: B}{(\gamma_{34,\times}-1)^2\over (\gamma_{\times}-1)^2}
181: \nu_{\rm m}^{\rm f}(t_{\rm \times})\propto n^{1/2},
182: \end{equation}
183: where ${\cal R}_{\rm B}$ is the ratio of the magnetic
184: field in the reverse emission region to that in the FS emission
185: region (Zhang, Kobayashi \& M\'{e}sz\'{a}ros 2003). Since at least for
186: some bursts (e.g. GRB990123 and GRB021211)
187: the RS emission region seems to be more strongly magnetized (e.g. Fan
188: et al. 2002; Zhang et al. 2003; Kumar \& Panaitescu 2003), here we
189: adopt two typical values, i.e. ${\cal R}_{\rm B}=5$ and 1, in the
190: calculations. There are two possibilities for a magnetized flow
191: (e.g. Fan et al. 2004). The central engine may
192: directly eject magnetized shells. Alternatively, the magnetic fields
193: generated in the internal shock phase may not be dissipated
194: significantly in a short period of time (e.g. Medvedev et al. 2005),
195: and they can get amplified again in the RS region. This second effect,
196: which has been ignored previously, should also play an important role
197: in calculating the afterglow re-brightening effect in refresh-shocks.
198:
199: Following Kobayashi \& Zhang (2003a) and Zhang et al. (2003), we have
200: \begin{equation}
201: \nu_{\rm c}^{\rm r}\approx {\cal R}_{\rm B}^{-3}\nu_{\rm c}^{\rm
202: f}\propto n^{-1},
203: \end{equation}
204: \begin{equation}
205: F_{\rm \nu, max}^{\rm r}(t_{\rm \times})\approx \eta {\cal R}_{\rm B}
206: F_{\rm \nu, max}^{\rm f}(t_{\rm \times})\propto n^{1/2}.
207: \end{equation}
208: Generally, the R-band flux satisfies $F_{\nu_{\rm R}}(t_\times)\approx
209: F_{\rm \nu,max}^{\rm r}(t_\times)[\nu_{\rm R}/\nu_{\rm m}^{\rm r}({\rm
210: t_\times})]^{\rm -(p-1)/2}\propto n^{\rm p+1\over 4}$. In the thin
211: shell case, the R-band RS flux is $F^r_{\rm \nu_{\rm
212: R}}\propto t_{\rm obs}^{\rm 2p}$ for $t_{\rm obs}<(1+z)t_{\rm
213: \times}$, and is $F^r_{\rm \nu_{\rm R}}\propto t_{\rm obs}^{-2}$ for
214: $t_{\rm obs}>(1+z)t_{\rm \times}$ (e.g., Sari \& Piran 1999; Kobayashi
215: 2000).
216:
217: If short GRBs are born in a stellar wind (for the collapsar model), for
218: the FS emission, the cooling frequency $\bar{\nu}_{\rm c}^{\rm f}$,
219: the typical synchrotron frequency $\bar{\nu}_{\rm m}^{\rm f}$ and the
220: maximum spectral flux $\bar{F}_{\rm \nu,max}^{\rm f}$ read (Chevalier
221: \& Li 2000)
222: \begin{equation}
223: \bar{\nu}_{\rm c}^{\rm f}=2\times 10^{13}{\rm Hz}~\epsilon_{\rm
224: B,-2}^{-3/2}E_{51}^{1/2}A_*^{-2}({2\over 1+z})t_{\rm d}^{1/2},
225: \label{Wind1}
226: \end{equation}
227: \begin{equation}
228: \bar{\nu}_{\rm m}^{\rm f}=4.5\times 10^{12}{\rm Hz}~\epsilon_{\rm
229: e,-0.5}^2\epsilon_{\rm B,-2}^{1/2}E_{51}^{1/2}({2\over 1+z})t_{\rm
230: d}^{-3/2},
231: \label{Wind2}
232: \end{equation}
233: \begin{equation}
234: \bar{F}_{\rm \nu,max}^{\rm f}\approx 3.8{\rm mJy}~\epsilon_{\rm
235: B,-2}^{1/2}E_{51}^{1/2}A_*D_{28.34}^{-2}({1+z\over 2})t_{\rm
236: d}^{-1/2},
237: \label{Wind3}
238: \end{equation}
239: where $A_*=(\dot{M}/10^{-5}M_\odot{\rm yr^{-1}})(v_{\rm w}/10^3{\rm
240: km~s^{-1}})^{-1}$, $\dot{M}$ is the mass loss rate of the progenitor,
241: and $v_{\rm w}$ is the wind velocity. Here the bar-parameters
242: represent the wind case.
243:
244: Equations (\ref{Wind1}-\ref{Wind3}) apply to an isotropic fireball, or
245: to a jet with opening angle $\theta_0$ when the bulk Lorentz factor
246: $\gamma > 1/\sqrt{3}\theta_0$, so that $\gamma\approx
247: 3.3E_{51}^{1/4}A_*^{-1/4}{t}_{\rm d}^{-1/4}$ is satisfied. For
248: $\gamma\leq 1/\sqrt{3}\theta_0$, if sideways expansion is significant,
249: the emission properties is similar to the ISM case (Sari et
250: al. 1999; Chevalier \& Li 2000). If sideways expansion is unimportant,
251: equations (\ref{Wind1}-\ref{Wind2}) still hold and equation
252: (\ref{Wind3}) should be replaced by $\bar{F}_{\rm \nu,max(J)}^{\rm
253: f}\approx \bar{F}_{\rm \nu,max}^{\rm f}(t_{\rm d}/\bar{t}_{\rm
254: 0,d})^{-1/2}$, where $\bar{t}_{\rm 0,d}$ is determined by
255: $3.3E_{51}^{1/4}A_*^{-1/4}\bar{t}_{\rm
256: 0,d}^{-1/4}=1/\sqrt{3}\theta_0$.
257:
258: In the wind case, the RS is usually relativistic (e.g., Chevalier \&
259: Li 2000). The resulting $t_{\times} \sim T_{90}$, and the optical
260: emission typically drops as $(t/t_\times)^{-3}$ for $t>t_\times$
261: (Kobayashi \& Zhang 2003b, Kumar \& Panaitescu 2000). For short
262: bursts, the duration when the reverse shock emission dominates is too
263: short for any observational interest. In this work, we do not include
264: the RS emission in the wind models.
265: Below we calculate the typical optical-band lightcurves for
266: short GRBs within different progenitor models.
267:
268: \subsection{Compact star merger model}
269:
270: The afterglows of short GRBs powered by mergers have been
271: investigated by Panaitescu et al. (2001) numerically.
272: Here we re-calculate the optical
273: afterglow lightcurve by also taking into account the RS emission.
274:
275: \begin{figure}
276: \epsscale{1.0}
277: \plotone{f1.eps}
278: \caption{The analytical R-band lightcurves of short
279: GRBs in the compact star merger model and the short emerging
280: model. The solid lines, dotted lines and the dashed line represent the
281: compact star merger model in the ISM environment, the short emerging
282: collapsar model in the ISM environment, and the same model in the wind
283: environment, respectively. For the first two models (the ISM models),
284: the reverse shock emission component was calculated for both ${\cal
285: R}_{\rm B}=5$ (the thin lines) and ${\cal R}_{\rm B}=1$ (the thick
286: lines). The thick dash-dotted line represents
287: the sensitivity of UVOT. For $t_{\rm obs}>5000{\rm s}$, the exposure
288: time of UVOT is assumed to be 1000s, while for $t_{\rm obs}<5000{\rm
289: s}$, it is assumed to be $t_{\rm
290: obs}/5$. Following parameters are adopted in the calculations:
291: $\eta=300$, $\epsilon_{\rm e}=0.3$, $\epsilon_{\rm B}=0.01$,
292: $p=2.3$, $z=1$, $D=2.2\times 10^{28}{\rm cm}$. In both the compact star
293: merger model and the short emerging model ISM model, it is assumed that the
294: outflow is jet-like with an opening angle $\simeq 0.1$ and an
295: isotropic energy $\simeq 10^{51}{\rm ergs}$. The ISM
296: number density is taken as $0.01{\rm cm^{-3}}$ and $1{\rm cm^{-3}}$,
297: respectively. For the short-merging wind model, the density is taken
298: as $n=3\times 10^{35}R^{-2}{\rm cm^{-3}}$.
299: For indicative purpose, we also plotted a template 1998bw-like
300: supernova R-band lightcurve at redshift $z=1$ (the line of plus
301: signs). }
302: \label{Light}
303: \end{figure}
304:
305:
306: The lightcurves for this model are plotted as solid lines in Figure
307: \ref{Light}. At the deceleration time [$\sim
308: 40(1+z){\rm s}$ after the burst trigger], the RS emission reaches
309: its peak, and the R band brightness is 20 mag for ${\cal R}_B=5$
310: (thin solid line) and
311: $z=1$. Swift UVOT has a sensitivity of 24 mag during 1000s exposure
312: time. Scaling down with time, the sensitivity should be 19 mag for 10s
313: exposure. Unless the event is much closer or $R_{\rm B}$ is larger,
314: the RS emission is likely to be below the UVOT sensitivity.
315: The FS emission is quite similar to the numerical calculation of
316: Panaitescu, et al. (2001). Because of a lower $n$ and a
317: smaller $E$, the R-band afterglow is much dimmer than that of typical
318: long GRBs, but it is still detectable by the UVOT for at least a few hours.
319: In the compact star merger scenario, the collimatation of the
320: outflow is quite uncertain. Here we adopt $\theta_0\sim 0.1$ as
321: suggested in numerical simulations (e.g. Aloy et al. 2004). As shown
322: in Figure \ref{Light}, the lightcurve break occurs too
323: late to be detected with the current telescope sensitivity.
324:
325: \subsection{Short emerging model}
326:
327: In the ``short emerging model'' (Zhang, Woosley \& MacFadyen 2003),
328: physical parameters (including the medium density $n$ and the jet opening
329: angle $\theta_{\rm 0}\simeq 0.1$) are generally similar to the familiar
330: long GRBs, except that the isotropic energy is smaller. This model has
331: received support from a recent comparison study of the spectral properties
332: of long and short GRBs (Ghirlanda, Ghisellini \& Celotti 2004).
333: The R-band lightcurves of this model are plotted as dotted lines in
334: Fig.\ref{Light} for ISM case, where the thin and thick lines are for ${\cal
335: R}_{\rm B}=5$ and ${\cal R}_{\rm B}=1$, respectively. Compared with
336: the compact star merger model, thanks to a larger $n$
337: ($F_{\rm \nu,max}^{\rm f}\propto n^{1/2}$ and $\nu_{\rm m}^{\rm
338: f}\propto n^0$ for $\gamma\geq (\sqrt{3}\theta_0)^{-1}$), the RS peak
339: flux is above the UVOT threshold, if ${\cal R}_{\rm B}$ is somewhat
340: larger than unity. The
341: RS emission peaks earlier (due to a smaller deceleration radius) so
342: that the RS peak may be missed if it is shorter than the slewing
343: time. In any case, the
344: $t_{\rm obs}^{-2}$ decaying component can be detected for ${\cal
345: R}_{\rm B}=5$ for a $z=1$ burst. In the wind case, for standard
346: parameters (e.g.
347: $n=3\times 10^{35}R^{-2}{\rm cm^{-3}}$ or $A_*=1$),
348: the resulting R-band lightcurve is very bright (see the thick dashed
349: line plotted in Fig. \ref{Light}), thanks to a relative denser medium
350: at $R<5.5\times 10^{17}{\rm cm}$.
351:
352: \subsection{Subjet model}
353:
354:
355: \begin{figure}
356: \epsscale{1.0}
357: \plotone{f2.eps}
358: \caption{The R-band lightcurves of short GRBs for the subjet
359: model. The typical lightcurve for the short emerging model is also
360: plotted for comparison. The upper panel is for the ISM case ($n=1{\rm
361: cm^{-3}}$), and the lower one is for the wind
362: case ($n=3.0\times10^{35}R^{-2}{\rm cm^{-3}}$). The thin lines are for
363: the subjet model.
364: The dotted, dash-dotted line, dashed line and solid lines represent
365: $\Delta \theta=0,~0.01,~0.02,~0.03$ respectively. For clarity, only
366: the forward shock emission is taken into account. Following parameters
367: are adopted. For the on-beam subjet, the jet opening angle is
368: $\theta_{\rm sub}=0.02$, and the isotropic energy is $10^{51}{\rm
369: ergs}$. For the Gaussian jet, the typical Gaussian angle is
370: $\theta_{\rm c}=0.08$, the maximum angle
371: is 0.3, and angle-dependent energy per solid angle reads
372: $\epsilon=(10^{53}/4\pi)\exp(-\theta^2/2\theta_{\rm c}^2)$. The
373: line-of-sight angle is $\theta_{\rm v}=0.26$ from the jet axis.
374: The thick solid line
375: is for the short emerging model calculated with the same code to
376: calculate the subjet model. The thick
377: dash-dotted line represents the sensitivity of UVOT. Other parameters
378: such as $\eta$, $\epsilon_{\rm e}$, $\epsilon_{\rm B}$, $p$ and $z$
379: are the same as those adopted to calculate Fig. \ref{Light}.
380: The supernova bump is also illustrated.}
381: \label{fig:Num}
382: \end{figure}
383:
384: In the subjet model (Yamazaki, Ioka \& Nakamura 2004), GRBs are
385: conjectured as being powered by many intrinsically similar subjets,
386: and the number of the subjets are
387: distributed with angle as a Gaussian function (Zhang \&
388: M\'{e}sz\'{a}ros 2002), i.e. $n\propto
389: \exp[-(\theta/\sqrt{2}\theta_{\rm c})^2]$, with the typical Gaussian
390: angle $\theta_{\rm c}\simeq 0.1$ (Zhang et al. 2004). If an observer
391: is far away from the jet axis and by chance is on the beam of one
392: subjet, one detects a short burst. The
393: global afterglow emission of this model could be then approximated by
394: that of a Gaussian structured jet superimposed on a uniform
395: subjet.
396: Here we consider two emission components, one on-beam uniform
397: less-energetic subjet with an opening angle $\theta_{\rm sub}\approx
398: 0.02$), and another stronger and wider Gaussian structured jet with
399: typical Gaussian angle $\theta_{\rm c}=0.08$ with the line-of-sight
400: angle $\theta_{\rm v}\simeq 3\theta_{\rm c}$ off-axis.
401: Since the Gaussian angular distribution is only of statistical sense in the
402: subjet model (Yamazaki et al. 2004), around the subjet there could be
403: a ``void'' where the emissivity is below the Gaussian jet model in
404: order to counterbalance the emissivity excess at the subjet. Here
405: we approximate this effect by adopting an annular void region of width
406: $\Delta\theta$ around the subjet axis (i.e. the emissivity is zero
407: in the range
408: from $\theta_{\rm sub}$ to $\theta_{\rm sub}+\Delta \theta$. In view
409: of the uncertainties, we calculate the lightcurves for $\Delta
410: \theta=0,~0.01,~0.02,~0.03$, respectively. Following Yamazaki et
411: al. (2004) we include a maximum Gaussian jet angle $\theta_{\rm j}
412: =0.3$ in the calculation.
413:
414: The afterglow lightcurves of structured jets have been
415: modeled by many authors (e.g., Wei \& Jin 2003; Kumar \& Granot 2003;
416: Granot \& Kumar 2003; Panaitescu \& Kumar 2003; Salmonson 2003; Rossi
417: et al. 2004).
418: Here we take the simple method
419: proposed by Wei \& Jin (2003), in which the sideways expansion of the
420: jet is ignored (see Kumar \& Granot 2003 for justification) but
421: the ``equal arriving surface'' effect is taken into account.
422: The jet evolution is quantified by $\gamma=(3\epsilon/n)^{1/2}(m_{\rm
423: p}c^2)^{-1/2}[ct/(1-\mu+1/16\gamma^2)]^{-3/2}$ for the ISM case, and
424: by $\gamma=(\epsilon/3\times 10^{35}A_*)^{1/2}(m_{\rm
425: p}c^2)^{-1/2}[ct/(1-\mu+1/8\gamma^2)]^{-1/2}$ for the wind
426: case\footnote{In the wind case, if we define $X \equiv
427: \epsilon/(3\times 10^{35}A_*m_{\rm p}c^3t)$, one has
428: $\gamma=[X(1-\mu)+\sqrt{X^2(1-\mu)^2+4X}]/2$, and the solutions could
429: be coasted into a simple form.}. Here
430: $\epsilon=(10^{53}/4\pi)\exp(-\theta^2/2\theta_{\rm
431: c}^2)$ is the energy per unit solid angle of the structured jet,
432: $\mu=\cos \Theta$, $\Theta$ is the angle between the moving direction
433: of an emitting unit and the line of
434: sight. The isotropic energy of the on-beam subjet is taken as
435: $10^{51}{\rm ergs}$.
436: The sideways expansion of the on-beam subjet is also ignored.
437: At any emission unit, the standard broken power-law
438: synchrotron spectrum (e.g. Sari et al. 1998) is adopted with $\delta
439: F^{\rm f}_{\rm \nu,max}\approx 3\sqrt{3}\Phi_{\rm p}(1+z)\delta N_{\rm
440: e}m_{\rm e}c^2\sigma_{\rm T} B/\{32 \pi^2eD^2[\gamma(1-\beta
441: \mu)]^3\}$ (Wijers \& Galama
442: 1999), where $\Phi_{\rm p}$ is a function of $p$ (for $p\simeq
443: 2.3$, $\Phi_{\rm p}\simeq 0.60$), $B$ is the magnetic field generated
444: at the shock front. In the ISM case, we take the total number of
445: electrons swept in the solid angle $d\Omega$ as $\delta N_{\rm
446: e}=d\Omega R^3 n/3$, where $R$ is the radius of the FS front. In the
447: wind case, $\delta N_{\rm e}=3.0\times 10^{35}Rd\Omega$ is adopted.
448:
449: The model lightcurves for the subjet model are plotted separately in
450: Fig.\ref{fig:Num}. The upper panel is the ISM case and the lower panel
451: is the wind
452: case. For a comparison, the lightcurve of short emerging model is
453: also plotted in each model (the thick solid line), which is similar to
454: the analytical result presented in Figure \ref{Light}. For the subjet
455: model, at the early times, the R-band emission is dominated by the
456: on-beam subjet. As the subjet is decelerated so that the Lorentz
457: factor is of order $\theta_{\rm sub}$, a very early jet break appears
458: (see Fig \ref{fig:Num} for detail).
459: On the other hand, the energetic Gaussian core component contributes
460: to the emission steadily, becomes progressively important at later
461: times, and dominates the afterglow level after thousands of seconds.
462: Because of the progressively important core contribution, the afterglow
463: decay in the subjet model is much slower than that in the short
464: merging model. Notice that the subjet model could be
465: different from the usual Gaussian jet model in which the angular
466: energy distribution is smooth (e.g. Kumar \& Granot 2003; Rossi et
467: al. 2004). The possible existence of the void around the subjet may lead to an
468: afterglow bump (see Figure \ref{fig:Num}). In fact, if $\Delta \theta$
469: is 0.1 or larger, the whole jet can be approximated as two
470: distinct components, i.e. a weak on-beam sub-jet and an off-beam
471: but more energetic uniform core since the result is insensitive to the
472: detailed structure in the core.
473: The bump can be then understood in terms of the off-beam orphan
474: afterglow models (e.g. Granot et al. 2002).
475: In our calculations the initial Lorentz factor across the whole jet is
476: assumed to be independent on the angle (Yamazaki et al. 2004).
477:
478: For both the short-emerging model and the subjet model, one may
479: expect a Type Ib/c supernova component (usually a red bump) showing up
480: a few weeks after the burst trigger, as has been detected in some long
481: GRBs. For illustrative purpose, we plot in Fig.\ref{Light} and
482: Fig. \ref{fig:Num} a template 1998bw-like supernova lightcurve at
483: $z=1$. The afterglows of short bursts are typically fainter than those
484: of the long ones, so the supernova signature should be easily
485: distinguishable, especially for the short emerging model.
486: For the subjet model, the contamination of the core component may
487: make the identification of the SN component more difficult. In any
488: case, if a flattening or bump is detected within weeks for a short GRB
489: afterglow, it would argue against the compact star merger model.
490:
491: \section{GRB 040924}
492:
493: GRB 040924 triggered the High Energy Transient Explorer 2 (HETE-2) on
494: 2004 September 24 at 11:52:11 UT (Fenimore et al. 2004). The burst
495: lasted $T_{50}\sim 1.2{\rm s}$, and the energy fluence was ${\cal
496: F}_{\gamma}\sim 7.7\times 10^{-6}{\rm ergs ~cm^{-2}}$ (Fenimore et
497: al. 2004; Golenetskii et al. 2004). The ratio of the fluence in the
498: 7-30 keV band and in the 30-400 keV band is about 0.6, so that the
499: burst is classified as an X-ray rich GRB. The burst redshift was
500: identified as $z=0.859$ (Wiersema et al. 2004).
501: The prompt localization of GRB 040924 by HETE-2
502: allowed follow-up observations of its afterglow at early times (Fox \&
503: Moon 2004; Li et al. 2004). Fox (2004) detected an optical transient
504: $\sim 16$ minutes after the trigger at the level of $m_{\rm R}\simeq
505: 18.0$mag. At the same position, Li et al. (2004) detected an optical
506: transient $\sim 26$ and $\sim 63$ minutes after the trigger at the
507: level of $m_{\rm R}\simeq 18.3$mag and $19.2$mag, respectively. Later
508: detections in K-band and R-band have been reported by many groups
509: (Terada \& Akiyama 2004; Terada, Akiyama \& Kawai 2004; Hu et
510: al. 2004; Fynbo et
511: al. 2004; Khamitov et al. 2004a, b, c). The radio observation provides
512: an upper limit of 0.12mJy at $\sim 15$ hours after the burst (van der
513: Horst 2004). Below we will compare the available data with the models,
514: aiming at constraining the burst environment and the possible
515: progenitor.
516:
517: \begin{figure}
518: \epsscale{1.0}
519: \plotone{f3.eps}
520: \caption{Modelling the R-band afterglow data of GRB
521: 040924. The identified burst redshift is $z=0.859$, and the total
522: fluence is ${\cal F}_{\gamma}=7.7\times 10^{-6}{\rm ergs}$ (Wiersema
523: et al. 2004). This gives $E_\gamma \simeq 1.5 \times 10^{52} {\rm
524: erg}$ assuming isotropic emission.
525: The data (marked by asterisk) are
526: taken from Fox (2004), Li et al. (2004), Hu et al. (2004) and
527: Kahamitov et al. (2004a, b, c). The solid line and dotted line
528: are the theoretical afterglow lightcurves of a
529: slow cooling fireball (or a jet with wide opening angle)
530: expanding into a low density ISM.
531: The parameters are $E=3\times 10^{52}{\rm ergs}$, $f_\gamma=2$,
532: $\epsilon_{\rm e}=0.1$, $\epsilon_{\rm B}=0.004$, $n=0.01{\rm
533: cm^{-3}}$ and $p=2.42$.
534: The solid and dotted lines are for ${\cal R}_{\rm
535: B}=3$ and ${\cal R}_{\rm B}=1$, respectively.}
536: \label{Mod}
537: \end{figure}
538:
539: \subsection{ISM case}
540: The constraint $F^{\rm f}_{\rm \nu,max}\geq 250{\rm \mu Jy}$ results in
541: $f_\gamma{\cal F}_{\gamma,-5.1}\epsilon_{\rm B,-2}^{1/2}n_{-2}^{1/2}\geq
542: 1.3$, where $f_\gamma\geq 1$ is the ratio of the afterglow energy to the
543: $\gamma-$ray energy. With $z=0.859$ and taking
544: $f_\gamma=2$, we can estimate $E\simeq 3\times 10^{52}{\rm ergs}$
545: within the standard cosmology. At the time
546: $t_{\rm R}\leq 945{\rm s}$, the typical frequency
547: of the FS emission crosses the observer frequency
548: (R-band, $\nu_{\rm obs}=4.6\times 10^{14}{\rm Hz}$). This results in
549: $0.12[{3(p-1)\over 13(p-2)}]^2(t_{\rm
550: R}/945 ~{\rm s})^{3/2}=E_{52.5}^{1/2}\epsilon_{\rm
551: B,-2}^{1/2}\epsilon_{\rm e,-0.5}^2(1+z)^{1/2}$. We then have the
552: following constraints
553: \begin{eqnarray}
554: &\epsilon_{\rm e}\leq 0.1[{3(p-1)\over 13(p-2)}]({t_{\rm R}\over
555: 945})^{3/4}E_{52.5}^{-1/4}(f_\gamma/2)^{1\over 2}{\cal F}_{\rm
556: \gamma,-5.1}^{1\over 2}n_{-2}^{1\over 4},\nonumber\\ &\epsilon_{\rm
557: B}\geq 4\times 10^{-3} (f_\gamma/2)^{-2}{\cal F_{\rm
558: \gamma,-5.1}}^{-2}n_{-2}^{-1}.\nonumber
559: \end{eqnarray}
560: The observed temporal decay slope is $\alpha_{\rm obs} \simeq -1.07$,
561: which gives $p=2.42$ in the standard afterglow model. The resulting
562: spectral index $\beta \simeq -0.71$ matches the observation
563: $\beta_{\rm obs}=0.61\pm 0.08$ (Silvey et al. 2004). Assuming $t_{\rm
564: R}\approx 945{\rm s}$ and $F^{\rm
565: f}_{\rm \nu,max}= 250{\rm \mu Jy}$ and $n=0.01{\rm cm^{-3}}$, one gets
566: $\epsilon_{\rm
567: e}\approx 0.1$ and $\epsilon_{\rm B}\approx 0.004$. The values
568: of the parameters $\epsilon_e$ and $\epsilon_B$ fall into the regime
569: inferred from
570: afterglow modeling of long bursts (Panaitescu \& Kumar 2002; Yost et
571: al. 2003). We note that if we take $n\sim 1{\rm cm^{-3}}$,
572: $\epsilon_{\rm B}\sim 10^{-5}$ is obtained. If the shock
573: parameters are more or less universal, our modeling suggests that a
574: low density ISM model is favored, which is consistent with the
575: pre-concept of the merger
576: model. In Fig.\ref{Mod}, we use our model lightcurves to fit the data.
577:
578: With parameters derived, $\nu_{\rm c}^{\rm f}$ is above
579: the optical energy band in all the observer time, which is consistent
580: with the observation (Silvey et al. 2004).
581:
582: \subsection{Wind case}
583: In the wind case, for $\beta_{\rm obs}\simeq 0.61 \pm 0.08$. With the temporal
584: index $\alpha_{\rm obs}\simeq -1.07$, $\bar{\nu}_{\rm m}^{\rm f}<\nu_{\rm
585: obs}<\bar{\nu}_{\rm c}^{\rm f}$ should be satisfied (e.g. Chevalier
586: \& Li 2000; see also the Tab. 1 of Zhang \& M\'{e}sz\'{a}ros 2004 for
587: a summary).
588:
589: At $945$s, the constraints of $\bar{\nu}_{\rm c}^{\rm f}>\nu_{\rm
590: obs}$, $\bar{\nu}_{\rm m}^{\rm f} \leq \nu_{\rm obs}$ and
591: $(\bar \nu_{\rm m}^{\rm f} / \nu_{\rm obs})^{\rm (p-1)/2} \bar{F}_{\rm
592: \nu,max}=250{\rm \mu Jy}$ yield
593: \begin{eqnarray}
594: &A_*<0.14(f_\gamma/2)^{1/4}\epsilon_{\rm B,-2}^{-3/4},\label{Wind4}\\
595: &\epsilon_{\rm e,-0.5}=0.09g^{1/2}(f_\gamma/2)^{-1/4}\epsilon_{\rm
596: B,-2}^{-1/4},\label{Wind5}\\
597: &A_*=5.9\times 10^{-4}g^{\rm -(p-1)/2}\epsilon_{\rm
598: B,-2}^{-1/2}(f_\gamma/2)^{-1/2}.\label{Wind6}
599: \end{eqnarray}
600: By taking
601: $\epsilon_{\rm B}\sim 10^{-3}$ and $f_\gamma=2$, we have
602: $\epsilon_{\rm e}\sim 0.05g^{1/2}$
603: and $A_*\sim 1.8\times 10^{-3}g^{\rm -(p-1)}<0.8$, where we have
604: defined $g=\bar \nu_{\rm m}^{\rm f} / \nu_{\rm obs}$. Therefore,
605: unless $\epsilon_{\rm e}$ is much smaller than the typical value
606: $0.1$, we get a very weak stellar wind $A_*\sim 10^{-3}$. A second
607: problem of the wind
608: model comes from the temporal index. For $\beta_{\rm obs}\simeq 0.61$
609: (Silvey et al. 2004), we have $p\simeq
610: 2.22$, which in turn results in $\alpha\simeq -1.4$. This is
611: significantly steeper than $\alpha_{\rm obs}$. We thus suggest that
612: the wind model is less favored.
613:
614: In summary, we suggest that the circumburst medium is preferably a
615: constant density ISM. If we believe that the shock parameters does not
616: vary significantly among bursts, the inferred $n$ is significantly
617: lower than that of the typical ISM, which coincides with the
618: pre-concept of the compact objects meger
619: model. No definite jet break is detected, so we do not know
620: the geometrically corrected $\gamma-$ray enengy. If GRB 040924 is
621: indeed powered by a merger event, no associated Ib/Ic supernova
622: signature (typically a red lightcurve bump with flux 1$\mu$Jy at
623: $z\sim 1$) is expected in a few weeks after the burst. The negative
624: detection of the supernova signature at the time when this work is
625: completed (two months after the burst trigger) is also consistent with
626: the compact star merger model.
627:
628: \section{Summary \& Discussion}
629:
630: We have modeled the typical optical afterglow lightcurves for short
631: bursts within the context of the leading progenitor models. Both the
632: forward and reverse shock emission components are considered.
633: With typical parameters, the early afterglows should be detectable
634: by the Swift UVOT, and a well-monitored lightcurve can help
635: to identify the progenitors of short bursts.
636:
637: The optical afterglow data collected so far for the recent bright
638: short burst GRB 040924 can be modeled well with an isotropic fireball
639: expanding into a low density medium with $n\sim 10^{-2}{\rm cm^{-3}}$.
640: The wind model is found to be less favored. The resulting parameters
641: are consistent with the pre-concept of the compact star merger model.
642: Other models such as a collapsar progenitor with low-density
643: environment, however, cannot be ruled out at this stage.
644: In principle, if GRB 040924 came from a collapsar, a lightcurve
645: flattening is expected within weeks resulting from either the
646: supernova component or the central core component for the subjet
647: model. The non-detection of such a feature so far presents a further
648: constraint on the collapsar model.
649:
650: GRB 040924 is a relatively soft event. It may not be a good
651: representative of the traditional short-hard bursts. Swift will
652: locate more short-hard bursts, and our analysis could be directly
653: utilized to discuss their nature.
654:
655: \acknowledgments YZF thanks D. M. Wei for helpful comments. We also
656: thank the anonymous referee for helpful suggestions. This work
657: is supported by NASA NNG04GD51G (for BZ), Eberly Research Funds of
658: Penn State and by the Center for Gravitational Wave Physics under
659: grants PHY-01-14375 (for SK), NASA AST 0098416 and NASA NAG5-13286
660: (for PM), and a NASA Swift GI (Cycle 1) program (for BZ, SK and PM).
661:
662: \begin{thebibliography}{99}
663: \bibitem[]{} Aloy, M. A., Janka, H. T., \& M\"{u}ller, E. 2004, A\&A,
664: submitted (astro-ph/0408291)
665: \bibitem[]{} Belczynski, K., Bulik, T., \& Kalogera, V. 2002, 571, L147
666: \bibitem[]{} Bloom, J. S., Sigurdsson, S., \& Pols, O. R. 1999, MNRAS,
667: 305, 763\bibitem[]{} Eichler, D., Livio, M., Piran, T., \& Schramm,
668: D. N. 1989, Nature, 340, 126
669: \bibitem[]{} Chevalier, R. A., \& Li, Z. Y. 2000, ApJ, 536, 195
670: \bibitem[]{} Fan, Y. Z., Dai, Z. G., Huang, Y. F., \& Lu, T. 2002,
671: Chin. J. Astron. Astrophys. 2, 449
672: \bibitem[]{} Fan, Y. Z., Wei, D. M., \& Wang, C. F. 2004, A\&A, 424, 477
673: \bibitem[]{} Fenimore, E. E., Ricker, G., Atteia, J-L., Kawai, N.,
674: Lamb, D., \& Woosley, S. 2004, GCN Circ. 2735
675: (http://gcn.gsfc.nasa.gov/gcn/gcn3/2735.gcn3)
676: \bibitem[]{} Fox, D. B. 2004, GCN Circ. 2741 (http://gcn.gsfc.nasa.gov
677: /gcn/gcn3/2741.gcn3)
678: \bibitem[]{} Fox, D. B., \& Moon, D. S. 2004, GCN Circ. 2734 (http://gcn.gsfc.
679: nasa.gov/gcn/gcn3/2734.gcn3)
680: \bibitem[]{} Fynbo, J. P. U., Hornstrup, A., Hjorth, J., Jensen,
681: B. L., \& Andersen, M. I. 2004, GCN Circ. 2747
682: (http://gcn.gsfc.nasa.gov/gcn/gcn3/2747.gcn3)
683: \bibitem[]{} Ghirlanda, G., Ghisellini, G., \& Celotti, A. 2004, A\&A,
684: 422, L55
685: \bibitem[]{} Golenetskii, S., Aptekar, R., Mazets, E., Pal'shin, V.,
686: \& Frederiks, D. 2004, GCN Circ. 2754 (http://gcn.gsfc.nasa.gov/gcn/gcn3
687:
688: /2754.gcn3)
689: \bibitem[]{} Granot, J., \& Kumar, P. 2003, ApJ, 591, 1086
690: \bibitem[]{} Granot, J., Panaitescu, A., Kumar, P., \& Woosley,
691: S. E. 2002, ApJ, 570, L61
692: \bibitem[]{} Hu, J. H., et al. 2004, GCN Circ. 2744 (http://gcn.gsfc.
693: nasa.gov/gcn/gcn3/2744.gcn3)
694: \bibitem[]{} Khamitov, I. et al. 2004a, GCN Circ. 2740
695: (http://gcn.gsfc.nasa.gov/gcn/gcn3/2740.gcn3)
696: \bibitem[]{} ------. 2004b, GCN Circ. 2749 (http://gcn.gsfc.nasa.gov/gcn/gcn3
697: /2749.gcn3)
698: \bibitem[]{} ------. 2004c, GCN Circ. 2752 (http://gcn.gsfc.nasa.gov/gcn/gcn3
699: /2752.gcn3)
700: \bibitem[]{} Kobayashi, S. 2000, ApJ, 545, 807
701: \bibitem[]{} Kobayashi, S., Piran, T. \& Sari, R. 1999, ApJ, 513, 669
702: \bibitem[]{} Kobayashi, S., \& Zhang, B. 2003a, ApJ, 582, L75
703: \bibitem[]{} ------. 2003b, ApJ, 597, 455
704: \bibitem[]{} Kumar, P. \& Granot, J. 2003, ApJ, 591, 1075
705: \bibitem[]{} Kumar, P. \& Panaitescu, A. 2000, ApJ, 541, L51
706: \bibitem[]{} ------. 2003, MNRAS, 346, 905
707: \bibitem[]{} Li, W., Filippenko, R., Chornock, R., \& Jha, S. 2004,
708: GCN Circ. 2748 (http://gcn.gsfc.nasa.gov/gcn/gcn3/2748.gcn3)
709: \bibitem[]{} Li, Z., Dai, Z. G., \& Lu, T. 2003, MNRAS, 345, 1236
710: \bibitem[]{} Lyutikov, M., \& Blandford, R. 2003 (astro-ph/0312347)
711: \bibitem[]{} Medvedev, M. V., et al. 2005, ApJ, 618, L75
712: \bibitem[]{} M\'{e}sz\'{a}ros, P. 2002, ARA\&A, 40, 137
713: \bibitem[]{} M\'{e}sz\'{a}ros, P., \& Rees, M. J. 1992, ApJ,
714: 397, 570
715: \bibitem[]{} Narayan, R., Paczy\'{n}ski, B., \& Piran, T. 1992, ApJ, 395, L83
716: \bibitem[]{} Paczy\'{n}ski, B. 1991, AcA, 41, 257
717: \bibitem[]{} Panaitescu, A., \& Kumar, P. 2002, ApJ, 571, 779
718: \bibitem[]{} ------. 2003, ApJ, 592, 390
719: \bibitem[]{} Panaitescu, A., Kumar, P., \& Narayan, P. 2001, ApJ, 561, L171
720: \bibitem[]{} Perna, R., \& Belczymski, K. 2002, ApJ, 570, 252
721: \bibitem[]{} Popham, R., Woosley, S. E., \& Fryer, C. 1999, ApJ, 518, 356
722: \bibitem[]{} Rees, M. J., \& M\'{e}sz\'{a}ros, P. 1992, MNRAS, 258, 41
723: \bibitem[]{} Rhoads, J. E. 1999, ApJ, 525, 737
724: \bibitem[]{} Rossi, E., Lazzati,D., Salmonson, J. D., \&
725: Ghisellini, G. 2004, MNRAS, 354, 86
726: \bibitem[]{} Rosswog, S., Ramirez-Ruiz, E., \& Davies, M. B. 2003,
727: MNRAS, 345, 1077
728: \bibitem[]{} Ruffert, M., et al. 1997, A\&A, 319, 122
729: \bibitem[]{} Salmonson, J. D. 2003, ApJ, 592, 1002
730: \bibitem[]{} Sari, R., \& Piran, T. 1999, ApJ, 517, L109
731: \bibitem[]{} Sari, R., Piran, T, \& Halpern, J. P. 1999, ApJ, 519, L17
732: \bibitem[]{} Sari, R., Piran, T, \& Narayan, R. 1998, ApJ, 497, L17
733: \bibitem[]{} Silvey, J. et al. 2004, GCN Circ. 2833 (http://gcn.gsfc.nasa.gov
734: /gcn/gcn3/2833.gcn3)
735: \bibitem[]{} Terada, H., \& Akiyama, M. 2004, GCN Circ. 2742
736: (http://gcn.gsfc.nasa.gov/gcn/gcn3/2742.gcn3)
737: \bibitem[]{} Terada, H., Akiyama, M., \& Kawai, N. 2004, GCN
738: Circ. 2750 (http://gcn.gsfc.nasa.gov/gcn/gcn3/2750.gcn3)
739: \bibitem[]{} Wang, X. Y., Cheng, K. S., \& Tam, P. H. 2004, ApJ, in press
740: \bibitem[]{} Wei, D. M., \& Jin, Z. P. 2003, A\&A, 400, 415
741: \bibitem[]{} Wijers, R. A. M. J., \& Galama, T. J. 1999, ApJ, 523, 177
742: \bibitem[]{} Wiersema, K., Starling, R. L. C., Rol, E.,
743: Vreeswijk, P., Wijers, R. A. M. J. 2004, GCN Circ. 2800
744: (http://gcn.gsfc.nasa.gov/gcn/gcn3/ 2800.gcn3)
745: \bibitem[]{} van der Horst, A. J., Rol, E., \& Wijers,
746: R. A. M. J. 2004, GCN Circ. 2746
747: (http://gcn.gsfc.nasa.gov/gcn/gcn3/2746.gcn3)
748: \bibitem[]{} Yamazaki, R., Ioka, K., \& Nakamura, T. 2004, ApJ, 607, L103
749: \bibitem[]{} Yost, S., Harrison,F. A., Sari, R., \& Frail,
750: D. A. 2003, ApJ, 597, 459
751: \bibitem[]{} Zhang, B., Dai, X., Lloyd-Ronning, N.M. \&
752: M\'{e}sz\'{a}ros, P. 2004, ApJ, 601, L119
753: \bibitem[]{} Zhang, B., Kobayshi, S., \& M\'{e}sz\'{a}ros, P. 2003, ApJ,
754: 595, 950
755: \bibitem[]{} Zhang, B., \& M\'{e}sz\'{a}ros, P. 2002, ApJ, 571, 876
756: \bibitem[]{} ------. 2004, Int. J. Mod. Phy. A., 19, 2385
757: \bibitem[]{} Zhang, W., Woosley, S. E., \& MacFadyen, A. I. 2003,
758: ApJ, 586, 356
759: \end{thebibliography}
760:
761: \end{document}
762:
763:
764:
765: