1: \documentstyle[emulateapj,psfig]{article}
2: %\documentclass[12pt,preprint]{aastex}
3: %\documentclass[12pt,referee]{aastex}
4: %\documentclass[useASM]{aastex}
5: \begin{document}
6:
7:
8: \title{The Peak Energy Distribution of the $\nu F_{\nu}$ Spectra and the Implications for the Jet Structure Models of Gamma-Ray Bursts
9: }
10:
11:
12: \author{E. W. Liang$^{1,2,3}$ and Z. G. Dai$^{1}$}
13: \affil{$^1$Department of Astronomy, Nanjing University, Nanjing 210093, China;
14: ewliang@nju.edu.cn.\\ $^2$Department of Physics, Guangxi University, Nanning 530004, China.\\
15: $^3$National Astronomical Observatories/Yunnan Observatory, Chinese Academy of Sciences, Kunming
16: 650011, China.}
17:
18:
19: \begin{abstract}
20: We study the peak energy ($E_{\rm{p}}$) distribution of the $\nu F_{\nu}$ spectra of gamma-ray
21: bursts (GRBs) and X-ray flashes (XRFs) with a sample of 57 bursts observed by {\em High Energy
22: Transient Explorer 2} ({\em HETE-2}) French Gamma Telescope and discuss its implications for the
23: jet structure models. Combining the observed $E_{\rm{p}}$ distribution of {\em HETE--2} GRBs/XRFs
24: with that of BATSE GRBs, we find that the observed $E_{\rm{p}}$ distribution of GRBs/XRFs is a
25: bimodal one with peaks of $\lesssim 30$ keV and $\sim 160-250$ keV. According to the
26: recently-discovered equivalent-isotropic energy-$E_{\rm{p}}$ relationship, such a bimodal
27: distribution implies a two-component structure of GRB/XRF jets. A simple simulation analysis shows
28: that this structured jet model does roughly reproduce a bimodal distribution with peaks of $\sim
29: 15$ and $\sim 200$ keV. We argue that future observations of the peak of $\sim 15$ keV in the
30: $E_{\rm{p}}$ distribution would be evidence supporting this model. {\em Swift}, which covers an
31: energy band of 0.2--150 keV, is expected to provide a key test for our results.
32: \end{abstract}
33: \keywords{gamma rays: bursts---gamma rays: observations---ISM: jets and outflows---methods: statistical} %Do NOT move this preamble from here!
34:
35: \section{Introduction} % Introduction goes below.
36: X-ray flashes (XRFs) have being gotten a lot of attention in the last 2 years (Heise et al. 2001;
37: Kippen et al. 2003). They are thought to be a lower energy extension of the known gamma-ray burst
38: (GRB) population, based on the fact that their spectral behaviors are similar to those of GRBs
39: (Kippen et al. 2003; Barraud et al. 2003; Sakamoto et al. 2004; Lamb et al. 2003a, b, 2004). The
40: nature of a narrow cluster of the observed $E_{\rm{p}}$ distribution of BATSE GRBs remains poorly
41: understood, which might be related to the jet structure of GRBs. XRFs broaden the energy coverage
42: of prompt GRB emission and may bring more signatures of the jet structure of GRBs (Lamb et al.
43: 2003a, b, 2004).
44:
45: The jet structure models are currently under heavy debate. Any model should present a unified
46: description for GRBs and XRFs. Two currently competing models are the structured jet model
47: (M\'{e}sz\'{a}ros, Rees \& Wijers 1998; Dai \& Gou 2001; Rossi, Lazzati, \& Rees 2002; Zhang \&
48: M\'{e}sz\'{a}ros 2002; Granot \& Kumar 2003; Kumar and Granot 2003; Panaitescu \& Kumar 2003; Wei
49: \& Jin 2003) and the uniform model (e.g., Rhoads 1999; Frail et al. 2001). Zhang et al. (2004) show
50: that the current GRB/XRF prompt emission/afterglow data can be described by a quasi--Gaussian-type
51: (or similar structure) structured jet with a typical opening angle of $\sim 6^{o}$ and with a
52: standard jet energy of $\sim 10^{51}$ ergs. Alternatively, based on the {\em High Energy Transient
53: Explorer 2} ({\em HETE-2}) observations, Lamb et al. (2003a, b, 2004) propose that the uniform jet
54: model can reasonably describe the unified scheme of GRBs/XRFs. Very recently, the two-component jet
55: model was advocated by Berger et al. (2003) based on observations of GRB 030329, which has two
56: different jet breaks in an early optical afterglow light curve (0.55 days, Price et al. 2003) and
57: in a late radio light curve (9.8 days). Millimeter observations of this burst further support the
58: two--component jet model (Sheth et al. 2003). Numerical calculations of such a model were performed
59: by Huang et al. (2004 ). This model suggests that a GRB/XRF jet has two components: a narrow,
60: highly relativistic one, and a wide, mildly relativistic one. When the line of sight of an observer
61: is within the narrow component, the observed burst is a typical GRB, but when the line of sight is
62: pointing to the wide--component, it is an XRF.
63:
64: A broad spectral energy distribution could constrain the jet structure models. A low peak energy of
65: the $\nu F_{\nu}$ spectrum ($E_{\rm{p}}<50$ keV) and weak gamma--ray fluxes ($F<0.2$ photons
66: cm$^{-2}$ s$^{-1}$, 50--300 keV energy range) distinguish XRFs from typical GRBs (Kippen et al.
67: 2003; Mochkovitch et al. 2003). It is well known that the observed $E_{\rm{p}}$ distribution of
68: BATSE GRBs is narrowly clustered. Does the observed $E_{\rm{p}}$ distribution of XRFs exhibit a
69: similar feature? In this Letter, we focus on this question. We analyze the observed $E_{\rm{p}}$
70: distribution with a sample of 57 bursts observed by {\em HETE-2}/ French Gamma Telescope (FREGATE).
71: Combining the observed $E_{\rm{p}}$ distribution of {\em HETE-2} GRBs/XRFs with that of BATSE GRBs,
72: we find that the observed $E_{\rm{p}}$ distribution of GRBs/XRFs is a bimodal one peaking at
73: $\lesssim 30$ keV and $\sim 160-250$ keV. With respect to this result, we suggest that the
74: two--component jet model is a reasonable candidate model for GRB/XRF jets. A simulation analysis
75: confirms this suggestion.
76:
77: \section{Distribution of $E_{\rm{p}}$}
78: We make a search for {\em HETE-2} GRBs/XRFs reported in literature and on the {\em HETE-2} Web
79: site\footnote{http://space.mit.edu/HETE/Bursts/}. All the bursts with $E_{\rm{p}}$ or fluences
80: ($S$) in the available energy bands 7--30 keV and 30--400 keV are included in our sample. We obtain
81: a sample that includes 57 bursts. Among them, 49 of the bursts are taken from Barraud et al.
82: (2003), Atteia (2003), Sakamoto et al. (2004), Lamb et al. (2003a, b, 2004), and the {\em HETE--2}
83: Web site. Their $E_{\rm{p}}$ values are derived from spectral fittings. Please note that the
84: $E_{\rm{p}}$ values of GRB 010923, 011216, and 021004 presented in Barraud et al. (2003) are
85: incorrect, and they are taken from Lamb et al. (2003a, b, 2004). For the other eight bursts, GRB
86: 030824, 030823, 030725, 030913, 030528, 030519, 030418, and 030416, only fluences in the energy
87: bands of 7--30 keV and 30--400 keV are available. For these bursts, we estimate their $E_{\rm{p}}$
88: by their spectral hardness ratios, which are defined as $R=S_{30-400\ {\rm{keV}}}/S_{7-30\
89: {\rm{keV}}}$. Since the spectra of GRBs/XRFs can be well fitted by the Band function (Band et al.
90: 1993) with similar spectral indices (Kippen et al. 2003; Barraud et al. 2003), their $E_{\rm{p}}$
91: should be proportional to $R$. A best fit to the data presented in Barraud et al. (2003) derives
92: $\log E_{\rm{p}}=(1.52\pm 0.05)+(0.92\pm 0.07)\log R$ with a linear coefficient of 0.93 and a
93: chance probability $p<0.0001$ ($N=32$, without considering GRB 010923, 011216, or 021004). We thus
94: estimate the $E_{\rm{p}}$ values of the above eight bursts by using this relation.
95:
96: We show the $E_{\rm{p}}$ distribution in a range of $\log E_{\rm{p}}/{\rm{keV}}=0.6-3.0$ with a
97: step of 0.23 for these bursts in Figure 1a. It is found that the distribution has three peaks at
98: 30, 160, and 450 keV. We note that the peaks of 160 and 450 keV seem to be embedded in one peak,
99: and the gap at $E_{\rm{p}}=275$ keV is likely to be fake. The spectral analysis for a bright BATSE
100: GRB sample by Preece et al. (2000) has shown that the $E_{\rm{p}}$ values are clustered at 100-1000
101: keV with a peak of $\sim 250$ keV (the dotted line in panel (a) of Figure 1). We thus suspect that
102: the peaks of 160 keV and 450 keV are likely to be embedded in one peak which is similar to that of
103: BATSE GRB sample. If the case really shows one peak, the $E_{\rm{p}}$ distributions observed by
104: {\em HETE--2} and by BATSE in the range of $100-1000$ keV should be consistent. We examine this
105: hypothesis by a Kolmogorov-Smirnoff (K-S) test (Press et al. 1997, p.617). The result of the K-S
106: test is described by a statistic of $P_{\rm{K-S}}$: a small value of $P_{\rm{K-S}}$ indicates a
107: significant difference between two distributions ($P_{\rm{K-S}}=1$ indicates that two distributions
108: are identical, and $P_{\rm{K-S}}<0.0001$ suggests that the consistency of two distributions should
109: be rejected; e.g., Bloom 2003). We obtain $P_{\rm{K-S}}=0.22$, indicating that the consistency of
110: the two distributions is acceptable. However, their difference is still quite significant. This
111: difference might be due to a strong sample selection effect in the BATSE GRB sample presented by
112: Preece et al. (2000), who considered only those bursts with total fluence $\geq 5\times 10^{-5}$
113: ergs cm$^{-2}$ or peak fluxes higher than 10 photons cm $^{-2}$ s$^{-1}$ in a 1.024 s timescale. To
114: avoid such a sample selection effect, we further compare the distributions of the hardness ratios
115: of {\em HETE-2} bursts and BATSE bursts in Figure 1b. In Figure 1b, the BATSE GRB sample includes
116: all of the long-duration bursts without any sample selection effect (1213 events, from BATSE
117: Current Catalog). A K--S test to the two distributions in the range of $\log R=0.3-1.5$ derives
118: $P_{\rm{KS}}=0.95$, strongly suggesting a consistency between the two distributions in that range.
119: Thus, we suggest that the $E_{\rm{p}}$ distribution in 100--1000 keV should form one sole peak,
120: centering at $\sim 160-250$ keV.
121:
122: The peak of $E_{\rm{p}}\sim 30$ keV or $R\sim 1$ seems to be a unique one. A sharp cutoff occurs on
123: its left side. This might be caused by the limit of {\em HETE-2}. Hence, we suggest that the
124: $E_{\rm{p}}$ distribution should exhibit another peak of an energy $\lesssim 30$ keV.
125:
126: Based on the above analysis, we propose that the $E_{\rm{p}}$ distribution of GRBs/XRFs is a
127: bimodal one, peaking at an energy $\lesssim 30$ keV and $\sim 160-250$ keV.
128:
129: \section{Implications for the Jet structure and Unified Models of GRBs/XRFs}
130: The observed bimodal distribution of $E_{\rm{p}}$ for GRBs/XRFs might strongly constrain the jet
131: structure models of GRBs/XRFs.
132:
133: From $E_{\rm{iso,52}}\simeq [E_{\rm{p,2}}(1+z)]^{2}$, where
134: $E_{\rm{iso,52}}=E_{\rm{iso}}/10^{52}\rm{\ ergs}$ and $E_{\rm{p,2}}=E_{\rm{p}}/10^2{\rm{\ keV}}$
135: (Amati et al. 2002; Lloyd-Ronning \& Ramirez-Ruiz 2002; Atteia 2003; Sakamoto et al. 2004; Lamb et
136: al. 2003a, b, 2004; Liang, Dai \&Wu 2004; Yonetoku et al. 2004), and
137: $E_{\rm{iso,52}}(1-\cos\theta_{\rm{j}})=0.133$, where $\theta_{\rm{j}}$ is the jet opening angle
138: (Frail et al. 2001; Panaitescu \& Kumar 2001; Piran et al. 2001; Bloom et al. 2003; Berger,
139: Kulkarni, \& Frail 2003), we can derive
140: \begin{equation}
141: \theta_{\rm{j}}={\rm{arccos}} \left \{ 1-\frac{0.133}{[E_{\rm{p,2}}(1+z)]^2}\right \}.
142: \end{equation}
143:
144: In the uniform jet model, one expects that both XRFs and GRBs should obey Eq. (1). However, this
145: relation cannot simply extend to any bursts with $E_{\rm{p}}(1+z)<35$ keV, because of the limit of
146: $\theta_{\rm{j}}<\pi/2$. The redshifts of the two extremely soft XRFs, 020903 and 030723, are 0.251
147: (Soderberg et al. 2003) and less than 2.1 (Fynbo et al. 2004), respectively; but their $E_{\rm{p}}$
148: values are less than $20$ keV. The two XRFs violate this relationship. In addition, the uniform jet
149: model may not accommodate the observed bimodal distribution of $E_{\rm{p}}$.
150:
151: A quasi-universal Gaussian-type jet model may also present a unified picture for GRBs/XRFs.
152: Lloyd-Ronning, Dai, \& Zhang (2003) found that this model can reproduce the relation of the
153: equivalent--isotropic energy to the viewing angle, and Zhang et al. (2004a) further showed that the
154: current GRB/XRF prompt emission/afterglow data can be described by this model (or similarly
155: structured jet) with a typical opening angle of $\sim 6^{\circ}$ and with a standard jet energy of
156: $\sim 10^{51}$ ergs. However, the observed bimodal distribution of $E_{\rm{p}}$ is difficult to be
157: explained by this model.
158:
159: According to the equivalent-isotropic energy--$E_{\rm{p}}$ relationship discovered recently by
160: Amati et al. (2002), the bimodal $E_{\rm{p}}$ distribution seems to imply a two-component structure
161: of GRB/XRF jets. To investigate whether or not this model can reproduce the observed bimodal
162: distribution of $E_{\rm{p}}$, we make a simple simulation analysis. We describe the energy per
163: solid angle of the two--component model by two Gaussian jets,
164:
165: \begin{equation}
166: \epsilon=\epsilon_0(e^{-\theta_{\rm{v}}^2/2\theta_{1}^2}+\mu e^{-\theta_{\rm{v}}^2/2\theta_2^2}),
167: \end{equation}
168: where $\theta_{\rm{v}}$ is the viewing angle measured from the jet axis, $\epsilon_0$ is the
169: maximum value of energy per solid angle, $\mu$ is the ratio of $E_{\rm{iso}}$ in the wide component
170: to narrow component, and $\theta_{1}$ and $\theta_{2}$ are characteristic angular widthes of the
171: narrow and wide components, respectively. Since $E_{\rm{p}} \propto \epsilon^{0.5}$, the observed
172: $E_{\rm{p}}$ should be given by
173:
174: \begin{equation}
175: E_{\rm{p}}=E_{\rm{p,\ 0}}(1+z)(e^{-\theta_{\rm{v}}^2/2\theta_{1}^2}+\mu
176: e^{-\theta_{\rm{v}}^2/2\theta_2^2})^{1/2}.
177: \end{equation}
178:
179: Similar to Lloyd-Ronning, Dai, \& Zhang (2003) and Zhang et al. (2004), we assume that the two
180: components are quasi-universal, where ``quasi" means that the parameters of this model have a
181: dispersion but are not invariable. We perform a simple Monte Carlo simulation analysis with the
182: distributions of these parameters. The probability of observing a GRB/XRF with $\theta_{\rm{v}}$ is
183: proportional to $\sin \theta_{\rm{v}}$. One can expect this probability to be random. Thus, we
184: assume that $\sin\theta_{\rm{v}}$ is uniformly distributed in the range of 0--1. The $E_{\rm{p,\
185: 0}}$ distribution should be mainly determined by a bright GRB sample. Since the observed
186: $E_{\rm{p}}$ for bright BATSE GRBs are narrowly clustered at $200-400$ keV and since the measured
187: redshift distribution is around $1$, we take the differential distribution of $E_{\rm{p,\ 0}}$ as
188: that of $E_{\rm{p}}$ for the bright GRBs, but centered at $\log E_{\rm{p,0}}=2.80$ (i.e.,
189: $E_{\rm{p,0}}=630$ keV), which is given by $w(\log E_{\rm{p,0}})=0.018 \exp \{-2[(\log
190: E_{\rm{p,0}}-2.80)^2]/0.45^2\}$, where the coefficient 0.018 is a normalized constant. We assume
191: that the redshift distribution is the same as the one of Bloom (2003), who assumed that the burst
192: rate as a function of redshift is proportional to the star formation rate, and who presented the
193: observed redshift distribution incorporating observational biases (model SF1 from Porciani \& Madau
194: 2001 is used in this work). We also restrict $z\leq4.5$ because the largest $z$ is 4.5 in our
195: present GRB sample. For $\theta_1$ and $\mu$, we cannot reasonably model their distributions with
196: the present data, and thus we simply estimate their values as follows. Since the mean value of the
197: jet opening angles of 16 GRBs presented in Bloom et al. (2003) is $\sim 0.15$ rad (without
198: considering the eight GRBs whose limits of jet opening angles are presented), we take $\theta_1\sim
199: 0.15$ rad. Based on the results shown in Figure 1, we have $\mu=E_{\rm{iso,XRF}}/E_{\rm{iso,GRB}}
200: \simeq 10^{-1.7}$. The $\theta_2$ is the most poorly understood among these parameters. We let it
201: be an adjustable variable with a limit of $\theta_2>\theta_1$. In our simulation analysis, we take
202: $\theta_2=0.32$ rad (see below).
203:
204: We simulate a sample of $10^5$ GRBs/XRFs. Our simulation analysis procedure is described as
205: follows. To derive a value of parameter $x$ for a given burst ($x$ is one of $E_{\rm{p,0}}$, $z$,
206: and $\theta_v$), we first derive the accumulative probability distributions of these parameters
207: $P(x)$ ($0<P(x)\leq 1$), then generate a random number $m$ ($0<m\leq 1$), and finally obtain the
208: value of $x$ from the inverse function of $P(x)=m$; i.e., $x=P^{-1}(m)$. The values of $\theta_1$
209: and $\mu$ are fixed at 0.15 rad and $10^{-1.7}$, respectively. The value of $\theta_2$ is an
210: adjustable variable with a limit of $\theta_2>\theta_1$. We find that $\theta_2=0.32$ rad can
211: roughly reproduce the $E_{\rm{p}}$ distribution shown in Figure 1. We calculate the $E_{\rm{p}}$
212: for each simulated GRB/XRF with the above parameters using Eq. (3). The $E_{\rm{p}}$ distribution
213: is shown in Figure 2. We find that the distribution is bimodal with peaks of $\sim 15$ and $\sim
214: 200$ keV and with a valley at $\sim 50$ keV. These results show that the two-Gaussian jet model can
215: roughly reproduce the bimodal distribution of the observed $E_{\rm{p}}$.
216:
217: In our simulation, we do not consider any instrument threshold setting. The energy bandpass of {\em
218: HETE--2}/FREGATE is 7--400 keV. From Figure 1, we find a sharp cutoff at $\log
219: E_{\rm{p}}/{\rm{keV}}=1.3$ (i.e., $E_{\rm{p}}=20$ keV), which is close to the lowest end of the
220: {\em HETE--2} energy bandpass. This $E_{\rm{p}}$ value might reflect the effective threshold of
221: {\em HETE--2}. We roughly estimate the ratio of observable GRBs to XRFs for {\em HETE--2} with this
222: threshold in our simulation analysis, and find that this ratio is about 2.2:1. This is in a good
223: agreement with {\em HETE--2}/FREGATE observations (39 {\em HETE--2} GRBs and 18 XRFs in the {\em
224: HETE--2}/FREGATE sample).
225:
226: \section{Conclusions and Discussion}
227: We have studied the observed $E_{\rm{p}}$ distribution of 57 {\em HETE-2}/FREGATE bursts, and
228: discuss its implications for the jet structure models. Combining the observed $E_{\rm{p}}$
229: distribution of {\em HETE-2} GRBs/XRFs with that of BATSE GRBs, we suggest that the observed
230: $E_{\rm{p}}$ distribution of GRBs/XRFs is bimodal with peaks of $\lesssim 30$ keV and $\sim
231: 160-250$ keV. According to the recently--discovered equivalent-isotropic energy--$E_{\rm{p}}$
232: relationship, we find that the bimodal distribution can be explained by the two--component model of
233: GRB/XRF jets. A simple simulation analysis shows that this structured jet model does roughly
234: reproduce the bimodal distribution with peaks of $\sim 15$ keV and $\sim 200$ keV.
235:
236: The peak of $\sim 15$ keV in the simulated $E_{\rm{p}}$ distribution is key evidence for the two--
237: component jet model. It is near the lowest end of the energy bandpass of {\em HETE-2}/FREGATE.
238: Fortunately, {\em HETE-2} provides a weak clue to this peak. A more sensitive instrument than {\em
239: HETE--2} with an energy bandpass $1-50$ keV is required to further confirm this peak. {\em Swift},
240: which covers an energy band of 0.2--150 keV (we mark this region in Figure 2 with diagonal
241: lines)\footnote{ http://swift.gsfc.nasa.gov/science/instruments/} is expected to provide a key test
242: for it.
243:
244: Simulations of the propagation and eruption of relativistic jets in massive Wolf-Rayet stars by
245: Zhang, Woosley, \& Heger (2004b) show that an erupting jet has a highly relativistic, strongly
246: collimated core, and a moderately relativistic, less energetic cocoon. The cocoon expands and
247: becomes visible at larger angles. The energy ratio of the cocoon to the core in their simulation is
248: about one order. From our simulation results, we find that it is $\sim
249: (E_{\rm{p,GRB}}/E_{\rm{p,XRF}})^2(\theta_{1}/\theta_2)^2\sim 40$, being roughly consistent with
250: their results. Their simulations seem to support the two-component jet model. We have noted that
251: the ability of the cocoon to cause an XRF depends sensitively on its Lorentz factor, which is
252: determined by the degree of mixing between the jet and envelope material. Matzner (2003) argued
253: that this mixing might be difficult to resolve in numerical simulations.
254:
255: A two-component jet was suggested to be universal for GRB/XRF phenomena in this Letter, based on
256: the multi-wavelength observations of GRB 0303029 (Berger et al. 2003; Sheth et al. 2003) and the
257: bimodal distribution of $E_{\rm {p}}$. It should be pointed out that other jet models such as
258: uniform jets and single-component-universal jets were proposed to explain numerous observations on
259: the afterglows and some correlations (e.g., Lamb et al. 2003b; Lloyd-Ronning et al. 2004;
260: Lloyd-Ronning \& Zhang 2004). Thus, one would expect strong evidence showing which jet model is
261: more reasonable.
262:
263: We are very grateful to the anonymous referee for his/her valuable suggestions. We also sincerely
264: thank Bing Zhang and Johan P.U. Fynbo for their helpful comments. These suggestions and comments
265: have enabled us to improve the manuscript greatly. This work is supported by the National Natural
266: Science Foundation of China (grants 10233010 and 10221001), the National 973 Project (NKBRSF
267: G19990754), the Natural Science Foundation of Yunnan (2001A0025Q), and the Research Foundation of
268: Guangxi University.
269:
270:
271:
272:
273: \begin{thebibliography}{}
274: % Please use the \harvarditem command to create references.
275: \bibitem{}Amati, L., et al. 2002, A\&A, 390, 81
276: \bibitem{}Atteia, J.-L. 2003, A\&A, 407, L1
277: \bibitem{}Band, D., et al. 1993, ApJ, 413, 281
278: \bibitem{}Barraud, C., et al. 2003, A\&A, 400, 1021
279: \bibitem{}Berger, E., Kulkarni, S. R., \& Frail, D. A. 2003, ApJ, 590,379
280: \bibitem{}Berger, E., et al. 2003, Nature, 426, 154
281: \bibitem{}Bloom, J., Frail, D. A., \& Kulkarni, S. R. 2003, ApJ, 588, 945
282: \bibitem{}Bloom, J. 2003, AJ, 125, 2865
283: \bibitem{}Dai, Z. G. \& Gou, L. J. 2001, ApJ, 552, 72
284: \bibitem{}Frail, D. A., et al. 2001, ApJ, 562, L55
285: \bibitem{}Fynbo, J. P. U. et al. 2004, ApJ, in press (astro-ph/0402240)
286: \bibitem{}Granot, J. \& Kumar, P. 2003, ApJ, 591, 1086
287: \bibitem{}Heise, J., in't Zand, J., Kippen, R. M., \& Woods, P. M. 2001, in Proc. 2nd Rome Workshop: Gamma-Ray Bursts in the Afterglow Era, eds. E. Costa, F. Fron--tera, J. Hjorth (Berlin: Springer--Verlag), 16
288: \bibitem{}Huang, Y. F., Wu, X. F., Dai, Z. G., Ma, H. T., \& Lu, T. 2004, ApJ, 605, 300
289: \bibitem{}Kippen, R. M., et al. 2003, in AIP Conf. Proc. 662, Gamma-Ray Burst and Afterglow Astronomy 2001, ed. G.
290: R. Ricker \& R. K. Vanderspek (Melville: AIP), 244
291: \bibitem{}Kumar, P. \& Granot, J. 2003, ApJ, 591, 1075
292: \bibitem{}Lamb, D. Q., Donaghy, T. Q., \& Graziani C. 2003a, preprint (astro--ph/0312504)
293: \bibitem{}Lamb, D. Q., Donaghy, T. Q., \& Graziani C. 2003b, preprint (astro--ph/0312634)
294: \bibitem{}Lamb, D. Q., et al. 2004, NewA Rev., 48, 423
295: \bibitem{}Lloyd--Ronning, N. M., \& Ramirez-Ruiz, E. 2002, ApJ, 576, 101
296: \bibitem{}Lloyd--Ronning, N. M., Dai X. Y., \& Zhang, B. 2004, ApJ, 601, 371
297: \bibitem{}Lloyd--Ronning, N. M., \& Zhang, B. 2004, ApJ, submitted (astro-ph/0404107)
298: \bibitem{}Liang, E. W., Dai, Z. G., \& Wu, X. F. 2004, ApJ, 606, L29
299: \bibitem{}Matzner, C. D. 2003, MNRAS, 345, 575
300: \bibitem{}M\'{e}sz\'{a}ros, P., Rees, M. J., \& Wijers, R. A. M. J. 1998, ApJ, 499, 301
301: \bibitem{}Mochkovitch, R., Daigne, F., Barraud, C., Atteia, J. 2003, astro-ph/0303289
302: \bibitem{}Panaitescu, A. \& Kumar, P. 2001, ApJ, 554, 667
303: \bibitem{}Panaitescu, A. \& Kumar, P. 2003, ApJ, 592, 390
304: \bibitem{}Press, W. H., Teukolsky, S. A., Vetterling, W. T. \& Flannerly, B. P 1997, in Numerical Recipes, (Cambrige: Cambrige University
305: Press)
306: \bibitem{}Price, P. A., et al. 2003, Nature, 423, 844
307: \bibitem{}Piran, T., Kumar, P., Panaitescu, A., \& Piro, L. 2001, ApJ, 560, L167
308: \bibitem{}Porciani, C. \& Madau, P. 2001, ApJ, 548, 522
309: \bibitem{}Preece, R. D., et al. 2000, ApJS, 126, 19
310: \bibitem{}Rhoads, J. E. 1999, ApJ, 525, 737
311: \bibitem{}Rossi, E., Lazzati, D., \& Rees, M. J. 2002, MNRAS, 332, 945
312: \bibitem{}Sakamoto, T., et al. 2004, ApJ, 602, 875
313: \bibitem{}Sheth, K., et al. 2003, ApJ, 595, L33
314: \bibitem{}Soderberg, A. M., et al. 2003, GCN 1554
315: \bibitem{}Wei, D. M., \& Jin, Z. P. 2003, A\&A, 400, 415
316: \bibitem{}Yonetoku, D., et al. 2003, ApJL, in press (astro-ph 0309217)
317: \bibitem{}Zhang, B. \& M\'{e}sz\'{a}ros P. 2002, ApJ, 571, 876
318: \bibitem{}Zhang, B., Dai X. Y., Lloyd-Ronning, N. M., \& M\'{e}sz\'{a}ros, P. 2004a, ApJ, 601, L119
319: \bibitem{}Zhang, W., Woosley, S. E., \& Heger, A. 2004b, ApJ, in press, astro-ph/0308389
320:
321: \end{thebibliography}
322:
323: \begin{figure}
324: \plotone{fig1.EPS} \caption{Observed $E_{\rm{p}}$ [panel (a)] and hardness ratio [panel (b) ]
325: distributions of {em HETE-2}/FREGATE GRBs/XRFs. In panel (a), the dashed line is the observed
326: $E_{\rm{p}}$ distribution of a bright BATSE GRB sample taken from Preece et al. (2000). In panel
327: (b), the dashed line is the observed hardness ratio distribution of all long-duration BATSE GRBs
328: without any sample selection effect (from BATSE Current Catalog).\label{fig1}}
329: \end{figure}
330:
331: \begin{figure}
332: \plotone{fig2.EPS} \caption{Simulated $E_{\rm{p}}$ distribution of the two-quasi-universal Gaussian
333: jet model. The diagonal line region is the energy band of {\em Swift}. \label{fig2}}
334: \end{figure}
335:
336: \end{document}
337: