1: %\documentstyle[emulateapj,psfig]{article}
2: \documentclass[12pt,preprint]{aastex}
3: %\documentclass[12pt,referee]{aastex}
4: %\documentclass[useASM]{aastex}
5: \begin{document}
6:
7: \title{The FWHM of local pulses and the corresponding power-law index of gamma-ray burst FRED pulses}
8:
9: \author{L.-W. Jia$^{1,3}$, Y.-P. Qin$^{1,2,4}$}
10: \affil{$^1$National Astronomical Observatories/Yunnan Observatory,
11: Chinese Academy of Sciences,\\P. O. Box 110, Kunming 650011, China\\
12: $^2$Physics science and technology institute, Guangxi University, Nanning, Guangxi 530004, P.R. China\\
13: $^3$The Graduate School of the Chinese Academy of Sciences\\
14: $^4$E-mail: ypqin@ynao.ac.cn; lwjia@ynao.ac.cn}
15:
16:
17: \begin{abstract}
18: The FWHM of gamma-ray burst (GRB) pulses is known to be related
19: with energy by a power-law. We wonder if the power-law index
20: $\alpha$ is related with the corresponding local pulse width
21: $FWHM_0$. Seven FRED (fast rise and exponential decay) pulse GRBs
22: are employed to study this issue, where six of them were
23: interpreted recently by the relativistic curvature effect (the
24: Doppler effect of fireballs) and the corresponding local pulses
25: were intensely studied. A regression analysis shows an
26: anti-correlation between $log \alpha$ and $log FWHM_0$ with a
27: slope of $-0.37\pm0.13$. This suggests that, for the class of the
28: GRB pulses which are consequences of the curvature effect, the
29: difference of the local pulse width might lead to the variation of
30: the power law index, where the smaller the width the larger the
31: value of $\alpha$. Since the number of sources employed in this
32: analysis is small, our result is only a preliminary one which
33: needs to be confirmed by larger samples.
34: \end{abstract}
35: \keywords{gamma rays: bursts}
36:
37: \section{Introduction}
38:
39: Gamma-ray bursts (GRBs) were discovered thirty-eight years ago by
40: chance, and now are recognized as the most luminous known objects
41: in the Universe (Fishman 2001). Since then, many observations of
42: the objects have been made, which have amassed a great deal of
43: information. Owing to their brief and random appearance in the
44: gamma-ray region, their study had become very difficult since
45: their discovery. Although the progress has been made in GRB
46: research, GRBs remain one of the most inexplicable astrophysical
47: phenomena observed today (Kouveliotou 1997).
48:
49: Temporal and spectral characteristics of prompt emission of
50: gamma-ray burst pulses have been intensely studied since they
51: could constrain the energizing and emission mechanisms (Ryde et
52: al. 2002). The correlation between GRB spectral and temporal
53: properties have been investigated by several research groups. It
54: was first noted by Norris et al. (1986) that GRB pulses exhibit a
55: hard-to-soft spectral evolution, and associated with it the pulses
56: are seen to be narrower at higher energies than they are at lower
57: bands which were confirmed by later works (see Fishman et al.
58: 1992; Link, Epstein, \& Priedhorsky 1993). By using the average
59: autocorrelation function and the average pulse width, Fenimore et
60: al. (1995) showed that the average pulse width has a power-law
61: dependence on energy with an index of about -0.4 (the range of it
62: is from -0.37 to -0.46, depending on how it is measured). This is
63: the first quantitative relationship between temporal and spectral
64: structure in gamma-ray bursts. Norris et al. (1996) found that
65: average raw pulse shape dependence on energy is approximately
66: power law, with an index of -0.40, consistent with the
67: autocorrelation analysis of Fenimore et al. (1995). Furthermore,
68: Nemiroff (2000) brought forward that over the energy range 100
69: keV-1 MeV in GRB 930214c (BATSE trigger 2193) the temporal scale
70: factors between a pulse measured at different energies are related
71: to that energy by a power law. The corresponding power-law indexes
72: found by Feroci et al. (2001) for GRB 990704 and by Piro et al.
73: (1998) for GRB 960720 are $-0.45$ and $-0.46\pm0.10$,
74: respectively. Costa (1999) also found that the power-law index for
75: GRB 960720 is $-0.46$, the same to Piro et al. (1998). The
76: spectral lag as a function of energy was examined for individual
77: pulses in GRBs (Norris et al. 2000), which confirmed the earlier
78: result of Fenimore et al. (1995). In a recent study (Crew et al.
79: 2003), the power-law relationship between the duration of GRB
80: 021211 and energy further confirmed the earlier result. The
81: anti-correlation between pulse widths and gamma-ray energy have
82: also been investigated by many other authors (e.g., Tavani 1997;
83: Wang et al. 2000; Beloborodov et al. 2000; Guidorzi et al. 2003;
84: Sakamoto et al. 2004; Dado et al. 2004).
85:
86: Norris et al. (1995) found an anti-correlation between $T_{90}$
87: and peak intensity, while a positive correlation between $T_{90}$
88: and total fluence was shown in Lee \& Petrosian (1997), and a
89: positive correlation between peak energy and variability was found
90: by Lloyd-Ronning \&\ Ramirez-Ruiz (2002). A correlation between
91: luminosity and variability for BATSE bursts with known redshifts
92: was revealed by Fenimore \& Ramirez-Ruiz (2000). Ramirez-Ruiz \&
93: Fenimore (2000) found a quantitative relationship between pulse
94: amplitude and pulse width: the smaller amplitude peaks tend to be
95: wider, with the pulse width following a power law with an index of
96: about -2.8 (the range of it is from -2.8 to -3.0, depending on how
97: it is measured). The anti-correlation between the pulse amplitude
98: and pulse width was also revealed by Lee et al. (2000). $T_{90}$
99: being correlated with peak heights (Lestrade 1994) and peak energy
100: being correlated with peak flux (Mallozzi et al. 1995) were other
101: reported relationships.
102:
103: It was suggested that, the most likely radiation progress in GRBs
104: is synchrotron emission in the standard fireball scenario (see
105: Katz 1994; Sari, Narayan, \& Piran 1996). The power-law dependence
106: has led to the suggestion that this effect could be attributed to
107: synchrotron radiation (see Piran 1999). Kazanas, Titarchuk, \& Hua
108: (1998) proposed that synchrotron cooling could well account for
109: the effect (see also Chiang 1998; Dermer 1998; and Wang et al.
110: 2000). Fenimore et al. (1995) showed that synchrotron emission can
111: give rise to the correlation $t_{syn}(E)\propto E^{-0.5}$ between
112: GRB spectral and temporal properties, which is consistent with the
113: observed correlation $\Delta \tau \propto E^{-0.45\pm 0.05}$.
114: Cohen et al. (1997) put forward that the power-law relationship
115: between pulse width and energy with the index of $-0.4$ is in
116: reasonable agreement with expectations for a population of
117: electrons losing energy by synchrotron radiation, for which an
118: exponent of $-1/2$ is predicted. It was suspected that a simple
119: relativistic mechanism might be at work in producing this
120: relationship (Nemiroff 2000). In deed, it was shown recently in
121: Qin et al. (2004; hereafter Paper I) and Qin et al. (2005) that
122: the Doppler effect of a relativistically expanding fireball
123: surface (the so-called relativistic curvature effect) could lead
124: to a power law relationship between the pulse width and energy for
125: FRED (fast rise and exponential decay) pulses, regardless the real
126: forms of the rest frame radiation and the local (or intrinsic)
127: pulse involved. The same effect was also observed by Shen et al.
128: (2005).
129:
130: In this paper, we investigate if local pulses are related with the
131: power law relationship (in other words, we wonder how the local
132: pulse width is related with the index of the power law observed).
133: In section 2, we choose several GRBs with each of them comprising
134: a single FRED pulse to calculate the corresponding data.
135: Relationship between the index and the local pulse width is
136: explored in section 3. Conclusions are presented in section 4.
137:
138: \section{Sources and data}
139:
140: To study how local pulses affect the index of the power law
141: between the pulse width and energy, we focus on FRED pulse bursts.
142: As revealed recently by many authors, the observed FRED structure
143: of pulses could be interpreted by the relativistic curvature
144: effect when the observed plasma moves relativistically towards us
145: and appears to be locally isotropic (see, e.g., Fenimore et al.
146: 1996; Ryde \& Petrosian 2002; Kocevski et al. 2003; Paper I; Shen
147: et al. 2005). If this interpretation is correct, FRED pulses would
148: form in nature a class identified by the GRB temporal structure.
149: In this way, it would not be great surprise to us if quantities
150: associated with the pulses are correlated with each other.
151:
152: As illustrated in Paper I, Shen et al. (2005), and Qin \& Lu
153: (2005), the local pulse width is essential to produce the observed
154: pulse shape due to the curvature effect. Accordingly, those FRED
155: pulses with their local pulses having been intensely studied
156: become our first choice. We find six bursts studied in Paper I
157: belonging to this kind. They are GRB 910721 ($\#563$), GRB 920925
158: ($\#1956$), GRB 930612 ($\#2387$), GRB 941026 ($\#3257$), GRB
159: 951019 ($\#3875$) and GRB 951102B ($\#3892$).
160:
161: Light curve data for which the background counts have been
162: subtracted are available in the BATSE website
163: (http://cossc.gsfc.nasa.gov/batse/batseburst/sixtyfour\_ms\\/bckgnd%
164: \_fits.html). The signal data are taken within the zone [$t_{\min
165: },t_{\max } $], where $t_{\max }-t_{\min }=2T_{90}$, and $t_{\min
166: }$ is at $T_{90}/2$ previous to the start of $T_{90}$.
167:
168: There might be many different methods to estimate the pulse width.
169: The theme of all possible methods is to find the central values of
170: the scattering data. In other words, one always manages to find
171: the real values of the data that are assumed to be get rid of the
172: chaos arising from the influence of the background as well as
173: other statistical errors. Owing to the fact that the light curve
174: function of Kocevski et al. (2003) (the KRL function; equation
175: [22] of Kocevski et al. 2003) could well describe the observed
176: light curves of FRED pulses (see also Qin \& Lu 2005), we simply
177: employ this function to fit the four channel light curves of the
178: six bursts, where parameters of the function associated with
179: different channels are allowed to be different for the same burst.
180: In order to allow the fitting curves shifting along the time axis
181: so that the time coordinate of the light curve data is unnecessary
182: to be resettled, we introduce an extra parameter $t_0$ to the KRL
183: function, where $t$ should be replaced by $t-t_0$ and $t_m$ should
184: be replaced by $t_m-t_0$. Thus, we have five free parameters
185: ($f_m$, $t_m$, $r$, $d$, $t_0$) in our fit, instead of four. The
186: widths of the four channel light curves are then estimated from
187: the corresponding fitting curves, where the errors are determined
188: by the uncertainties of the fitting parameters via the error
189: transfer formula.
190:
191: We perform the fit with the software of ORIGIN, where the fitting
192: parameters as well as their uncertainties are available.
193: Illustrated in Figure. 1 are the fits to the four channel light
194: curves of GRB 951019 ($\#3875$). For GRB 941026 ($\#3257$) and GRB
195: 951102B ($\#3892$), the widths in channel 4 are not available
196: since the signal in that channel is too weak to be detected. The
197: estimated values of the FWHM of the observed light curves of the
198: six bursts calculated with the fitting curves (determined by the
199: fitting parameters) are listed in Table 1.
200:
201: Assuming that the widths of pulses are related with energies by a
202: power law, we calculate the indexes with the estimated values of
203: the observed pulse widths of the six sources. The results are
204: presented in Table 2.
205:
206: There are several local pulses discussed in Paper I, some of which
207: are as follows:
208:
209: 1. The local pulse with an exponential decay
210: \begin{equation}
211: \widetilde{I}(\tau _\theta )=I_0\exp (-\frac{\tau _\theta -\tau
212: _{\theta ,\min }}\sigma )\qquad(\tau _{\theta ,\min }\leq \tau
213: _\theta )
214: \end{equation}
215:
216: 2. The local pulse with a power-law rise and a power-law decay
217: \begin{equation}
218: \widetilde{I}(\tau _\theta )=I_0\{
219: \begin{array}{c}
220: (\frac{\tau _\theta -\tau _{\theta ,\min }}{\tau _{\theta ,0}-\tau
221: _{\theta ,\min }})^\mu \qquad(\tau _{\theta ,\min }\leq
222: \tau _\theta \leq \tau _{\theta ,0}) \\
223: (1-\frac{\tau _\theta -\tau _{\theta ,0}}{\tau _{\theta ,\max
224: }-\tau _{\theta ,0}})^\mu \qquad(\tau _{\theta ,0}<\tau _\theta
225: \leq \tau _{\theta ,\max })
226: \end{array}
227: \end{equation}
228:
229: 3. The local pulse with a power-law rise
230: \begin{equation}
231: \widetilde{I}(\tau _\theta )=I_0(\frac{\tau _\theta -\tau _{\theta ,\min }}{%
232: \tau _{\theta ,\max }-\tau _{\theta ,\min }})^\mu \qquad(\tau
233: _{\theta ,\min }\leq \tau _\theta \leq \tau _{\theta ,\max })
234: \end{equation}
235:
236: In Paper I, local pulse (1) was adopted to account for the light
237: curves of GRB 910721 and GRB 930612 when the curvature effect was
238: considered, for GRB 941026 and GRB 951102B local pulse (2) with
239: $\mu =1$ was taken, while or GRB 920925 and GRB 951019 local pulse
240: (3) with $\mu =1$ was assumed. After smoothing the signal data,
241: they got a very good fit to these sources (see the $\chi^2$ values
242: listed in Table 2 of Paper I), which suggests that the assumption
243: that the observed light curves could arise from the local pulses
244: adopted when taking into account the curvature effect is
245: acceptable.
246:
247: According to the local pulse parameters listed in Table 2 of Paper
248: I, we get from equations (1)-(3) the widths of the corresponding
249: local pulses (note that $\tau _{\theta ,\min }=0$ was adopted in
250: Paper I), which are listed in Table 2 in this paper as well.
251:
252: \section{Relationship analysis}
253:
254: Relation between the index of the power law, $\alpha$, and the
255: FWHM of the local pulses, $FWHM_0$, is displayed in Figure 2. A
256: linear correlation between $log \alpha$ and $log FWHM_0$ could be
257: observed.
258:
259: We wonder if sources other than those selected in Paper I are in
260: agreement with this trend. As a FRED pulse source, GRB 930214c
261: ($\#2193$) was previously intensely studied (Nemiroff 2000). We
262: include this burst in our study. As done in the case of the six
263: bursts, we once more employ the KRL function to fit the four
264: channel light curves of this source, and in the same way,
265: parameters of the function associated with different channels are
266: allowed to be different. Also, the fit is performed with the
267: software of ORIGIN. The widths of the 4 channels of this burst are
268: estimated with the fitting parameters, which are listed in Table 1
269: as well. From these widthes we get the power law index of this
270: source under the assumption that the widths are related with
271: energies by a power law. The estimated value of the index is
272: presented in Table 2.
273:
274: To obtain the local pulse width of this burst, we follow what were
275: done by Qin et al. in Paper I for the six GRBs. One can find the
276: details of the analysis in the mentioned paper, which are omitted
277: in the following. Briefly stating, we fit the count rate of the
278: third channel of GRB 930214c ($\#2193$) with equation (21) of
279: Paper I, where local pulse (1) in this paper (which is local pulse
280: [83] in Paper I) is adopted. Relations (e.g., $t=t_1\tau+t_0$, see
281: Paper I for a detailed explanation) and functions (e.g., DB3) and
282: the corresponding parameters taken for the fit are exactly those
283: adopted in Paper I in the case of GRB 910721. The fit yields:
284: $\sigma=1.70$, $\chi^2_\nu=0.416$ for the data smoothed with DB3
285: wavelet in the level of the first-class decomposition,
286: $\chi^2_\nu=0.819$ for the data without smoothing, and other free
287: parameters (they are not related to the local pulse width). (Owing
288: to the limited space provided, the figure showing the fit is
289: omitted.) We find for GRB 930214c ($\#2193$) that the reduced
290: $\chi^2$ associated with the fit is reasonable, which suggests
291: that the light curve of this burst could indeed be accounted for
292: by the relativistic curvature effect.
293:
294: The data point of ($\alpha$, $FWHM_0$) for GRB 930214c ($\#2193$)
295: is also plotted in Figure 2, which is in agreement with the trend
296: mentioned above (see Figure 2).
297:
298: A linear correlation analysis of the data of the seven bursts
299: yields: $log \alpha=(-0.43\pm0.06)+(-0.38\pm0.11)log FWHM_0 $
300: ($r=-0.84$, $N=7$). However, it should be noticed that the number
301: of the sources concerned is small. In this case, the result of the
302: correlation analysis might obviously depend on some lonely located
303: data points (see GRB 920925 and GRB 930214c in Figure 2).
304: According to Isobe et al. (1990) and Feigelson \& Babu (1992), the
305: true regression coefficient uncertainty in samples of small size
306: would be underestimated when the usual standard formulas are
307: applied. Thus, resampling procedures such as the jackknife or
308: bootstrap should be used to evaluate regression uncertainties in
309: these cases. We thus try to use the bootstrap method to estimate
310: the regression coefficient uncertainties. Applying the bootstrap
311: error analysis we get indeed a larger slope uncertainty: $log
312: \alpha=(-0.43\pm0.08)+(-0.37\pm0.13)log FWHM_0 $. This is what we
313: should hold.
314:
315: \section{Conclusions}
316:
317: In this paper, we investigate the relationship between the
318: power-law index $\alpha$ and the FWHM of local pulses, $FWHM_0$,
319: of seven FRED pulse GRBs. Our analysis shows that there exists a
320: linear relationship with a slope of $-0.37\pm0.13$ between $log
321: \alpha$ and $log FWHM_0$ for the bursts. This suggests that
322: different widths of local pulses could lead to different values of
323: the power law indexes, with the larger the former the smaller
324: absolute value the latter, for FRED pulse bursts (at least for
325: those which were previously interpreted by the relativistic
326: curvature effect). If this relationship could be confirmed, the
327: distribution of the local pulse width would be an important factor
328: that leads to the variation of the index observed in GRB samples
329: (this might likely be true if the sample contains only FRED
330: pulses).
331:
332: Of the seven GRBs, local pulses of six were previously intensely
333: studied and that of the other one is explored in this paper. As
334: the number of the sources involved is small, the result is only a
335: preliminary one, which is not at all conclusive in terms of
336: statistics. However, a trend in the relationship is explicitly
337: illustrated in our analysis, although the analysis is qualitative
338: rather than quantitative. A large sample of FRED pulses is thus
339: required to check statistically if this conclusion could hold.
340:
341: Say frankly, the cause of this relationship is currently unclear.
342: Since the seven bursts studied here are single FRED pulse sources
343: which were assumed to suffer from the relativistic curvature
344: effect, we suspect that it might be this effect that gives birth
345: to the relationship. A theoretical analysis on this issue is
346: necessary.
347:
348: Besides the curvature effect, there might be other factors that
349: can affect the value of the power law index. One would be the
350: variation of the rest frame emission mechanism, which was revealed
351: in Qin et al. (2005). For example, different rest frame spectra or
352: different speeds of the rest frame spectral softening could lead
353: to different values of the power law index. This also requires a
354: further investigation.
355:
356: \acknowledgments Our thanks are given to Dr. Robert Nemiroff for
357: providing us helpful suggestions which make the paper
358: significantly improved. This work was supported by the Special
359: Funds for Major State Basic Research Projects (``973'') and
360: National Natural Science Foundation of China (No. 10273019).
361:
362: \begin{thebibliography}{Beloborodov, Stern, \& Svensson 2000}
363: \bibitem[Beloborodov, Stern, \& Svensson 2000]{Be00} Beloborodov, A. M.,
364: Stern, B. E., \& Svensson, R. 2000, ApJ, 535, 158
365:
366: \bibitem[Chiang 1998]{Ch98} Chiang, J., 1998, ApJ, 508, 752
367:
368: \bibitem[Cohen et al. 1997]{Co97} Cohen, E., Katz, J. I., Piran, T., \&
369: Sari, R. 1997, ApJ, 488, 330
370:
371: \bibitem[Costa 1999]{Co99} Costa, E. 1999, Nuclear Physics B (Proc. Suppl.), 69/1-3,
372: 646
373:
374: \bibitem[Crew et al. 2003]{Cr03} Crew, G. B., Lamb, D. Q., Ricker, G. R.,
375: Atteia, J.-L., Kawai, N., et al. 2003, ApJ, 599, 387
376:
377: \bibitem[Dado, Dar, \& R\'{u}jula 2004]{Da04} Dado, S., Dar, A., \& R\'{u}%
378: jjula, A. D. 2004, A\&A, 422, 381
379:
380: \bibitem[Dermer 1998]{De98} Dermer, C. D. 1998, ApJ, 501, L157
381:
382: \bibitem[Feigelson \& Babu 1992]{Fe92} Feigelson, E. D., Babu, G. J. 1992, ApJ, 397, 55
383:
384: \bibitem[Fenimore et al. 1995]{Fe95} Fenimore, E. E., In't Zand, J. J. M.,
385: Norris, J. P., Bonnell, J. T., \& Nemiroff, R. J. 1995, ApJ, 448,
386: L101
387:
388: \bibitem[Fenimore et al. 1996]{Fe96} Fenimore, E. E., Madras, C. D.,
389: \& Nayakshin, S. 1996, ApJ, 473, 998
390:
391: \bibitem[Fenimore \& Ramirez-Ruiz 2000]{FR00} Fenimore, E. E., \&
392: Ramirez-Ruiz, E. 2000, ApJ, submitted [astro-ph/0004176]
393:
394: \bibitem[Feroci et al. 2001]{Fe01} Feroci, M., Antonelli, L. A., Soffitta,
395: P., In't Zand, J. J. M., Amati, L., et al. 2001, A\&A, 378, 441
396:
397: \bibitem[Fishman et al. 1992]{Fi92} Fishman, G., et al. 1992, in Gamma-Ray
398: Bursts: Huntsville, 1991, ed. W. S. Paciesas \& G. J. Fishman (New
399: York: AIP), 13
400:
401: \bibitem[Fishman 2001]{Fi01} Fishman, G. J. 2001, sgrb.conf, 9F
402:
403: \bibitem[Guidorzi et al. 2003]{Gu03} Guidorzi, C., Frontera, F., Montanari,
404: E., Amati, L., Antonelli, L. A., et al. 2003, A\&A, 401, 491
405:
406: \bibitem[Isobe et al. 1990]{Is90} Isobe, T., Feigelson, E. D., Akritas, M. G., Babu, G. J.
407: 1990, ApJ, 364, 104
408:
409: \bibitem[Katz 1994]{Ka94} Katz, J. I. 1994, ApJ, 422, 248
410:
411: \bibitem[Kazanas, Titarchuk, \& Hua 1998]{Ka98} Kazanas, D., Titarchuk, L.
412: G., \& Hua, X.-M. 1998, ApJ, 493, 708
413:
414: \bibitem[Kocevski, Ryde, \& Liang 2003]{Ko03} Kocevski, D., Ryde, F., \&
415: Liang, E. 2003, ApJ, 596, 389
416:
417: \bibitem[Kouveliotou 1997]{Ko97} Kouveliotou, C. 1997, AAS, 191, 3601
418:
419: \bibitem[Lee \& Petrosian 1997]{LP97} Lee, T. T., \& Petrosian, V. 1997,
420: ApJ, 474, L37
421:
422: \bibitem[Lee, Bloom, \& Petrosian 2000]{Le00} Lee, A., Bloom, E. D., \&
423: Petrosian, V. 2000, ApJS, 131, 21
424:
425: \bibitem[Lestrade 1994]{Le94} Lestrade, J. P. 1994, ApJ, 429, L5
426:
427: \bibitem[Link, Epstein, \& Priedhorsky 1993]{Li93} Link, B., Epstein, R.
428: I., \& Priedhorsky, W. C. 1993, ApJ, 408, L81
429:
430: \bibitem[Lloyd-Ronning \& Ramirez-Ruiz 2002]{LR02} Lloyd-Ronning, N. M., \&
431: Ramirez-Ruiz, E. 2002, ApJ, 576, 101
432:
433: \bibitem[Mallozzi et al. 1995]{Ma95} Mallozzi, R. S., Paciesas, W. S.,
434: Pendleton, G. N., Briggs, M. S., Preece, R. D., Meegan, C. A., \&
435: Fishman, G. J. 1995, ApJ, 454, 597
436:
437: \bibitem[Nemiroff 2000]{Ne00} Nemiroff, R. J., 2000, ApJ, 544, 805
438:
439: \bibitem[Norris et al. 1986]{No86} Norris, Share, G. H., Messina, D. C. et al. 1986, ApJ, 301, 213
440:
441: \bibitem[Norris et al. 1995]{No95} Norris, J. P., Bonnell, J. T., Nemiroff,
442: R. J., Scargle, J. D., Kouveliotou, C., Paciesas, W. S., Meegan,
443: C. A.,\& Fishman, G. J. 1995, ApJ, 439, 542
444:
445: \bibitem[Norris et al. 1996]{No96} Norris, J. P., Nemiroff, R. J., Bonnell,
446: J. T., Scargle, J. D., Kouvelitou, C., et al. 1996, ApJ, 459, 393
447:
448: \bibitem[Norris, Marani, \& Bonnell 2000]{No00} Norris, J. P., Marani, G.
449: F., \& Bonnell, J. T. 2000, ApJ, 534, 248
450:
451: \bibitem[Piran 1999]{Pi99} Piran, T. 1999, Phys. Rep., 314, 575
452:
453: \bibitem[Piro et al. 1998]{Pi98} Piro, L., Heise, J., Jager, R., Costa, E.,
454: Frontera, F., et al. 1998, A\&A, 329, 906
455:
456: \bibitem[Qin et al. 2004]{Qi04} Qin, Y.-P., Zhang, Z.-B., Zhang, F.-W., \&
457: Cui, X.-H. 2004, ApJ, 617, 439 (Paper I)
458:
459: \bibitem[Qin et al. 2005]{Qi05a} Qin, Y.-P., Dong, Y.-M., Lu, R.-J., Zhang, B.-B., Jia, L.-W. 2005, ApJ,
460: in press [astro-ph/0411365 v2]
461:
462: \bibitem[Qin \& Lu 2005]{Qi05b} Qin, Y.-P., Lu, R.-J. 2005, MNRAS,
463: in press [astro-ph/0504406]
464:
465: \bibitem[Ramirez-Ruiz \& Fenimore 2000]{RF00} Ramirez-Ruiz, E., \&
466: Fenimore, E. E. 2000, ApJ, 539, 712
467:
468: \bibitem[Ryde \& Petrosian 2002]{RP02} Ryde, F., \& Petrosian, V. 2002,
469: ApJ, 578, 290
470:
471: \bibitem[Sakamoto et al. 2004]{Sa04} Sakamoto, T., Lamb, D. Q., Graziani,
472: C., Donaghy, T. Q., Suzuki, M., et al. 2004, ApJ, 602, 875
473:
474: \bibitem[Sari, Narayan, \& Piran 1996]{Sa96} Sari, R., Narayan, R., \&
475: Piran, T. 1996, ApJ, 473, 204
476:
477: \bibitem[Shen et al. 2005]{Sh05} Shen, R.-F., Song, L.-M., Li, Z. 2005, MNRAS, in
478: press [astro-ph/0505276]
479:
480:
481: \bibitem[Tavani 1997]{Ta97} Tavani, M., 1997, ApJ, 480, 351
482:
483: \bibitem[Wang et al. 2000]{Wa00} Wang, J. C., Cen, X. F., Qian, T. L., Xu,
484: J., \& Wang, C. Y. 2000, ApJ, 532, 267
485: \end{thebibliography}
486:
487: \clearpage
488:
489: \begin{figure}[tbp]
490: \includegraphics[angle=0,scale=1.5]{f1.eps}
491: \caption{Illustration of the fit with the KRL function to the four
492: channel light curves of GRB 951019.}
493: \end{figure}
494:
495: \clearpage
496:
497: \begin{figure}[tbp]
498: \includegraphics[angle=0,scale=1.5]{f2.eps}
499: \caption{Plot of the power law index versus the FWHM of local
500: pulses. The solid line is the linear regression line of the data.
501: }
502: \end{figure}
503:
504: \clearpage
505:
506: \begin{table*}
507: \begin{flushleft}
508:
509:
510: Table 1. The FWHM of the observed light curves of various channels
511: estimated with the KRL function for GRB 910721($\#563$), GRB
512: 920925($\#1956$), GRB 930214c($\#2193$), GRB 930612($\#2387$), GRB
513: 941026($\#3257$), GRB 951019($\#3875$) and GRB
514: 951102B($\#3892$), respectively.\\[0pt]
515:
516: \centering
517: \vspace{0.2cm}
518: \tabcolsep0.02in
519: \begin{tabular}[htb]{cccccccc} \hline \hline
520:
521: GRB&trigger&&channel&&$FWHM$(s)&&$\sigma_{FWHM} $ \\
522: \hline
523: 910721 &$\#$563 && 1 && 13.16 && 3.49 \\
524: & && 2 && 9.55 && 0.75 \\
525: & && 3 && 5.35 && 0.36 \\
526: & && 4 && 2.30 && 0.77 \\
527: \hline
528: 920925 &$\#$1956 && 1 && 5.55 && 0.76 \\
529: & && 2 && 4.68 && 0.73 \\
530: & && 3 && 4.23 && 0.82 \\
531: & && 4 && 5.91 && 4.13 \\
532: \hline
533: 930214c&$\#$2193 && 1 && 36.65 && 7.74 \\
534: & && 2 && 43.69 && 2.00 \\
535: & && 3 && 26.79 && 0.75 \\
536: & && 4 && 13.93 && 2.40 \\
537: \hline
538: 930612 &$\#$2387 && 1 && 18.78 && 0.75 \\
539: & && 2 && 15.39 && 0.39 \\
540: & && 3 && 12.39 && 0.31 \\
541: & && 4 && 6.85 && 2.53 \\
542: \hline
543: 941026 &$\#$3257 && 1 && 23.86 && 3.61 \\
544: & && 2 && 17.07 && 1.03 \\
545: & && 3 && 8.98 && 0.25 \\
546: \hline
547: 951019 &$\#$3875 && 1 && 1.58 && 0.22 \\
548: & && 2 && 1.33 && 0.20 \\
549: & && 3 && 0.77 && 0.08 \\
550: & && 4 && 0.59 && 0.58 \\
551: \hline
552: 951102B&$\#$3892 && 1 && 2.99 && 0.38 \\
553: & && 2 && 1.84 && 0.49 \\
554: & && 3 && 1.47 && 0.30 \\
555: \hline \hline
556: \end{tabular}
557: \end{flushleft}
558: \end{table*}
559:
560: \clearpage
561:
562: \begin{table*}[tbp]
563: \begin{flushleft}
564:
565: Table 2. Estimated values of the power law index of the
566: 7 bursts and the FWHM of the corresponding local pulses. \\[0pt]
567:
568: \centering
569:
570: \vspace{0.1cm}
571: \tabcolsep0.02in
572: \begin{tabular}[htb]{cccccccccc} \hline \hline
573:
574: GRB&trigger&&$\alpha$&&&&$\sigma_{\alpha} $&&$FWHM_0$ \\
575: \hline
576: 910721 & $\#$563 && 0.77&&&& 0.12&& 0.14 \\
577: \hline
578: 920925 & $\#$1956 && 0.17&&&& 0.16&& 2.97 \\
579: \hline
580: 930214c & $\#$2193 && 0.62&&&& 0.07&& 1.18 \\
581: \hline
582: 930612 & $\#$2387 && 0.29&&&& 0.03&&1.46 \\
583: \hline
584: 941026 & $\#$3257 && 0.81&&&& 0.08&& 0.12 \\
585: \hline
586: 951019 & $\#$3875 && 0.50&&&& 0.12&& 0.26 \\
587: \hline
588: 951102B & $\#$3892 && 0.49&&&& 0.18&& 0.41 \\
589: \hline \hline
590: \end{tabular}
591: \end{flushleft}
592: \end{table*}
593:
594:
595: \label{lastpage}
596:
597: \end{document}
598: