1: % v6. (07/14/07)
2: % Re-draft based on the complete simulations
3: % Hans' check for English
4: %
5: % v7. (05/07/08)
6: % simulation has been expanded for Ep grids
7: % T. Sakamoto (referee's comment)
8: %
9: % 11/03/08
10: % 3rd round of revision
11:
12: %% define
13: \def\eiso{E_{\rm iso}}
14: \def\egamma{E_{\gamma}}
15: \def\ep{E_{\rm peak}}
16: \def\epo{E^{\rm obs}_{\rm peak}}
17: \def\eps{E^{\rm src}_{\rm peak}}
18: \def\lpiso{L^{\rm peak}_{\rm iso}}
19:
20: \documentclass[12pt,preprint]{aastex}
21: \usepackage{color}
22: %% manuscript produces a one-column, double-spaced document:
23:
24: %% \documentclass[manuscript]{aastex}
25:
26: %% preprint2 produces a double-column, single-spaced document:
27:
28: %% \documentclass[preprint2]{aastex}
29:
30: %% Sometimes a paper's abstract is too long to fit on the
31: %% title page in preprint2 mode. When that is the case,
32: %% use the longabstract style option.
33:
34: %% \documentclass[preprint2,longabstract]{aastex}
35:
36: %% If you want to create your own macros, you can do so
37: %% using \newcommand. Your macros should appear before
38: %% the \begin{document} command.
39: %%
40: %% If you are submitting to a journal that translates manuscripts
41: %% into SGML, you need to follow certain guidelines when preparing
42: %% your macros. See the AASTeX v5.x Author Guide
43: %% for information.
44:
45: \newcommand{\vdag}{(v)^\dagger}
46: \newcommand{\myemail}{takanori@milkyway.gsfc.nasa.gov}
47:
48: %% You can insert a short comment on the title page using the command below.
49:
50: \slugcomment{Not to appear in Nonlearned J., 45.}
51:
52: %% If you wish, you may supply running head information, although
53: %% this information may be modified by the editorial offices.
54: %% The left head contains a list of authors,
55: %% usually a maximum of three (otherwise use et al.). The right
56: %% head is a modified title of up to roughly 44 characters.
57: %% Running heads will not print in the manuscript style.
58:
59: \shorttitle{$\Gamma$-$\ep$ relation}
60: \shortauthors{Sakamoto et al.}
61:
62: %% This is the end of the preamble. Indicate the beginning of the
63: %% paper itself with \begin{document}.
64:
65: \begin{document}
66:
67: %% LaTeX will automatically break titles if they run longer than
68: %% one line. However, you may use \\ to force a line break if
69: %% you desire.
70:
71: \title{$\ep$ estimator for Gamma-Ray Bursts Observed by the
72: {\it Swift} Burst Alert Telescope}
73:
74: %% Use \author, \affil, and the \and command to format
75: %% author and affiliation information.
76: %% Note that \email has replaced the old \authoremail command
77: %% from AASTeX v4.0. You can use \email to mark an email address
78: %% anywhere in the paper, not just in the front matter.
79: %% As in the title, use \\ to force line breaks.
80:
81: \author{T. Sakamoto\altaffilmark{1,2,3},
82: G. Sato\altaffilmark{8},
83: L. Barbier\altaffilmark{3},
84: S. D. Barthelmy\altaffilmark{3},
85: J. R. Cummings\altaffilmark{1,2,3},
86: E. E. Fenimore\altaffilmark{5},
87: N. Gehrels\altaffilmark{3},
88: D. Hullinger\altaffilmark{11},
89: H. A. Krimm\altaffilmark{1,7,3},
90: D. Q. Lamb\altaffilmark{9},
91: C. B. Markwardt\altaffilmark{1,6,3},
92: D. M. Palmer\altaffilmark{5},
93: A. M. Parsons\altaffilmark{3},
94: M. Stamatikos\altaffilmark{4,3},
95: J. Tueller\altaffilmark{3},
96: T. N. Ukwatta\altaffilmark{12,3}
97: }
98:
99: \altaffiltext{1}{Center for Research and Exploration in Space Science
100: and Technology (CRESST), NASA Goddard Space Flight Center, Greenbelt, MD
101: 20771}
102: \altaffiltext{2}{Joint Center for Astrophysics, University of Maryland,
103: Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250}
104: \altaffiltext{3}{NASA Goddard Space Flight Center, Greenbelt, MD 20771}
105: \altaffiltext{4}{Oak Ridge Associated Universities, P.O. Box 117,
106: Oak Ridge, Tennessee 37831.}
107: \altaffiltext{5}{Los Alamos National Laboratory, P.O. Box 1663, Los
108: Alamos, NM, 87545.}
109: \altaffiltext{6}{Department of Astronomy, University of Maryland,
110: College Park, MD 20742.}
111: \altaffiltext{7}{Universities Space Research Association, 10211 Wincopin
112: Circle, Suite 500, Columbia, MD 21044.}
113: \altaffiltext{8}{Institute of Space and Astronautical Science,
114: JAXA, Kanagawa 229-8510, Japan.}
115: \altaffiltext{9}{Department of Astronomy and Astrophysics, University of
116: Chicago, Chicago, IL, 60637.}
117: \altaffiltext{10}{Joint Center for Astrophysics, University of Maryland, Baltimore County,
118: 1000 Hilltop Circle, Baltimore, MD 21250}
119: \altaffiltext{11}{Moxtek, Inc., 452 West 1260 North, Orem, UT 84057}
120: \altaffiltext{12}{Department of Physics, The George Washington University,
121: Washington, D.C. 20052}
122:
123: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
124: %% style, abstract will output a Received/Accepted line after the
125: %% title and affiliation information. No date will appear since the author
126: %% does not have this information. The dates will be filled in by the
127: %% editorial office after submission.
128:
129: \begin{abstract}
130: We report a correlation based on a spectral simulation study
131: of the prompt emission spectra of
132: gamma-ray bursts (GRBs) detected by the {\it Swift} Burst Alert
133: Telescope (BAT).
134: The correlation is between the $\ep$ energy, which is
135: the peak energy in the $\nu$F$_{\nu}$ spectrum, and the photon index
136: ($\Gamma$) derived from a
137: simple power-law model. The $\ep$ - $\Gamma$ relation, assuming
138: the typical smoothly broken power-law spectrum of GRBs, is
139: %$\log \ep = 3.025 - 0.690\,\Gamma$ ($1.3 \leq \Gamma \leq 2.3$).
140: $\log \ep = 3.258 - 0.829\,\Gamma$ ($1.3 \leq \Gamma \leq 2.3$).
141: We take into account not only a range of $\ep$
142: energies and fluences, but also distributions for both the
143: low-energy photon index and the high-energy photon index in the smoothly
144: broken power-law model. The distribution of burst durations in the BAT
145: GRB sample is also included in the simulation.
146: Our correlation is consistent with the
147: index observed by BAT and $\ep$ measured by the BAT, and by other
148: GRB instruments.
149: Since about 85\% of GRBs observed by the BAT are acceptably fit
150: with the simple power-law model because of the relatively narrow energy
151: range of the BAT, this relationship can be used to estimate $\ep$
152: when it is located within the BAT energy range.
153: \end{abstract}
154:
155: %% Keywords should appear after the \end{abstract} command. The uncommented
156: %% example has been keyed in ApJ style. See the instructions to authors
157: %% for the journal to which you are submitting your paper to determine
158: %% what keyword punctuation is appropriate.
159:
160: %% Authors who wish to have the most important objects in their paper
161: %% linked in the electronic edition to a data center may do so in the
162: %% subject header. Objects should be in the appropriate "individual"
163: %% headers (e.g. quasars: individual, stars: individual, etc.) with the
164: %% additional provision that the total number of headers, including each
165: %% individual object, not exceed six. The \objectname{} macro, and its
166: %% alias \object{}, is used to mark each object. The macro takes the object
167: %% name as its primary argument. This name will appear in the paper
168: %% and serve as the link's anchor in the electronic edition if the name
169: %% is recognized by the data centers. The macro also takes an optional
170: %% argument in parentheses in cases where the data center identification
171: %% differs from what is to be printed in the paper.
172:
173: \keywords{gamma rays: bursts}
174:
175: %% Plan
176: %% 1. Introduction
177: %% - Importance of Epeak (Epeak-Eiso, Epeak-Egamma, ...)
178: %% - Difficulty of Epeak in Swift/BAT data
179: %% 2.1 BAT data analysis
180: %% - heasoft, selection, catalog, ...
181: %% 2.2 Data selection
182: %% - chi2 > 6,
183: %% - table 1,
184: %% - fluence distribution of our sample
185: %% 2.3 Epeak vs. photon index
186: %% Approximate formula Epeak<100 keV
187: %% 2.4 Spectral simulation,
188: %% Goro's results
189: %% 3. Discussion/Conclusion
190:
191: \section{Introduction}
192:
193: One of the fundamental characteristics of the prompt emission of
194: gamma-ray bursts (GRB) is $\ep$, which is the peak energy
195: in the $\nu$F$_{\nu}$ spectrum. According to {\it Beppo}SAX and {\it HETE-2}
196: observations, $\ep$ for GRBs is widely spread from a few keV
197: to the MeV range as a single distribution \citep{kippen_xrf_astroph,sakamoto2005}.
198: This broad single $\ep$ distribution strengthens the argument
199: that these bursts arise from the same origin.
200: Based on this observational evidence, there are
201: several works which try to understand a unified picture of GRBs. For
202: instance,
203: {\it the off-axis jet model} \citep{yamazaki2004,toma2005},
204: {\it the structured jet model} \citep{rossi2002,zhang2002,zhang2004},
205: and
206: {\it the variable jet opening angle model} \citep{lamb2005} are the
207: popular unified jet models. On the other hand, there are
208: theoretical
209: models to explain the broad $\ep$ distribution in the frame work of
210: the internal shock model
211: \citep{mes2002,mochkovitch2003,barraud2005} and the external shock model
212: \citep{dermer1999,huang2002,dermer2003}.
213:
214: There are several important empirical relationships proposed based
215: on the $\ep$ energy. One of the most cited relationships is the
216: correlation between $\ep$ in the GRB rest frame ($\eps$) and
217: the isotropic radiated energy ($\eiso$), the so called the $\eps$-$\eiso$ (Amati)
218: relation \citep{amati2002,amati2003}. Since this relation is extended
219: down to X-ray flashes \citep{sakamoto2004,sakamoto2006}, the dynamic range
220: of this relation is $\sim$3 orders of magnitude in $\eps$ and $\sim$5 order
221: of magnitude in $\eiso$. The second correlation is between the $\eps$
222: energy and the collimation-corrected energy ($\egamma$), the so called
223: $\eps$-$\egamma$ (Ghirlanda) relation \citep{ghirlanda2004}.
224: According to \citet{ghirlanda2004}, this relation has much tighter
225: correlation than the $\eps$-$\eiso$ relation. \citet{liang2005}
226: investigated a similar relationship, but without using $\egamma$
227: which is heavily dependent on the calculation of the jet opening angle.
228: They found a good correlation between $\eps$, $\eiso$, and the
229: achromatic break time in the afterglow light curve (t$_{jet}$).
230: The third relationship is between $\eps$ and the isotropic peak luminosity
231: ($\lpiso$), the so called the $\eps$-$\lpiso$ (Yonetoku) relation
232: \citep{yonetoku2004}. The latest fourth relationship is between
233: $\lpiso$, $\eps$, and the time scale of the brightest 45 per cent of
234: the background subtracted counts in the light curve of the prompt
235: emission \citep{firmani2006}.
236: If these relationships are valid, they must be related to the fundamental
237: physics of GRBs. Thus, $\eps$ energy provides us fruitful knowledge
238: about the characteristics of the prompt emission of GRBs. Furthermore,
239: knowing the $\epo$ energy is crucial to calculating the bolometric fluence
240: which reflects the
241: total radiated energy in the prompt emission.
242:
243: After the launch of $Swift$ \citep{gehrels2004} in 2004, the Burst Alert
244: Telescope (BAT; \citet{barthelmy2005}) has observed about 100
245: GRBs per year. In about half of the GRBs,
246: the $\ep$ energies are very likely to be within the BAT energy range
247: \citep{sakamoto2007}. However, due to the relatively narrow
248: energy band of the BAT (15-150 keV in the background subtracted spectrum using the
249: mask modulation), the BAT has a difficulty in determining $\epo$. Our
250: purpose of this study is to find a way to estimate $\epo$ when it lies
251: within the BAT energy range.
252:
253: Here, we report a good correlation between the photon power-law index derived
254: from a simple power-law model and $\ep$ based on the spectral simulation study.
255: We use a sample of 31 long BAT GRBs that are well fitted with the power-law times
256: exponential cutoff model, and also 26 GRBs observed by other GRB instruments
257: concurrent with the BAT to confirm our correlation. Our correlation
258: provides an estimate for $\ep$ from the photon index in a simple power-law
259: fit at the range from 1.3 to 2.3. We also calculated the 1$\sigma$ confidence
260: level of the estimated $\ep$ of our correlation.
261:
262: \section{BAT spectral simulation}
263: Because of the systematic difference in the spectral parameters based on
264: the assumption of the spectral model \citep{band1993}, we decided to perform
265: the simulations for two typical GRB spectral models as input spectra:
266: the smoothly broken power-law model (Band function; \citet{band1993})
267: \footnote{dN/dE = K$_{1}$E$^{\Gamma_{1}} \exp[-\rm{E}(2+\Gamma_{1})/\ep]$ if
268: E $< (\Gamma_{1} - \Gamma_{2}) \ep/(2+\Gamma_{1})$ and
269: dN/dE = K$_{2}$E$^{\Gamma_{2}}$ if $E \geq
270: (\Gamma_{1} - \Gamma_{2}) \ep/(2+\Gamma_{1})$}
271: and a power-law times exponential cutoff
272: model\footnote{dN/dE $\sim$ E$^{\alpha} \exp(-(2+\alpha)\,\rm{E}/\ep$)} (CPL) model.
273: % More explicitly
274: We fit the low-energy photon index, $\alpha$, and high-energy photon index,
275: $\beta$, of 124 samples of the Band function fit (``BAND'' in their notation)
276: in Table 9 of \citet{kaneko2006} by the normal distribution. We obtained
277: $\alpha$ of $-0.87$ with $\sigma$ of 0.33 and $\beta$ of $-2.36$ with $\sigma$
278: of 0.31.
279: %% 2nd ref comment
280: Note that we are not excluding the case of $\beta > -2$ in our simulations because
281: two reports \citep[e.g.,][]{rsato2005,kaneko2006} show fits with $\beta > -2$ in
282: both time-averaged and time-resolved burst spectra.
283: %%% 3rd ref comment
284: However, the fraction of simulated spectra with $\beta > -2$ is
285: only 13\% of the total.
286: %%
287: Similarly, for a CPL model, we fit the low-energy photon index,
288: $\alpha_{\rm CPL}$, for the sample in Table 9 of \citet{kaneko2006} (``COMP''
289: in their notation; 67 samples) by the normal distribution. We found $\alpha_{\rm CPL}$ of $-1.11$
290: with $\sigma$ of 0.30 (see Figure \ref{sim_input_paramters}). These $\alpha$,
291: $\beta$, and $\alpha_{\rm CPL}$ distributions are used in our spectral simulation.
292:
293: %use the best fit normal
294: %distribution of the low-energy photon index, $\alpha_{\rm CPL}$, centering
295: %on $-1.11$ with $\sigma$ of 0.30 similarly based on the BATSE spectral parameters
296: %(see Figure \ref{sim_input_paramters}).
297: In our simulations, $\ep$ varies from 1.4 keV to 1210 keV in a logarithmic
298: scale. The 15-150 keV fluence varies from 5 $\times$ 10$^{-8}$ to 5 $\times$
299: 10$^{-5}$ ergs cm$^{-2}$ in a logarithmic scale. The fluence range is determined
300: based on the BAT observations (BAT1 catalog; \citet{sakamoto2007}).
301: The simulation used 20 values for fluence and 70 values for $\ep$.
302: The exposure time of the spectrum is the best fit log-normal
303: distribution of the BAT T$_{100}$ duration\footnote{The duration includes
304: from 0 to 100\% of the GRB fluence.} reported in the BAT1 catalog\footnote{The duration
305: between tstart and tstop time of the fluence table.}
306: (See the bottom panel of Figure \ref{sim_input_paramters}).
307: The normalization of the input spectrum is calculated to be the
308: input fluence value. The spectral simulations are performed 1000 times
309: for each grid point. The background is included in the simulation using the
310: spectrum created from the event data of the false BAT trigger 180931. Since
311: the background is subtracted using the mask modulation, the
312: exposure time of the background spectrum is set as the same as the
313: duration of the foreground spectrum. Four incident angles,
314: on-axis (0$^{\circ}$), 15$^{\circ}$, 30$^{\circ}$, and 50$^{\circ}$ off-axis,
315: are simulated independently. The simulated spectra
316: are fitted from 14 keV to 150 keV with a simple power-law
317: model\footnote{dN/dE $\sim$ $E^{-\Gamma}$} (PL), a CPL, and
318: the Band function. {\tt Xspec} 11.3.2 was used in both creating and fitting
319: the simulated spectra.
320:
321: Figure \ref{fig:sim_ep_fluence} shows the numbers of the simulated spectra
322: which have $\Delta \chi^{2}$ (
323: $\Delta \chi^{2}$ $\equiv \chi^{2}_{\rm PL} - \chi^{2}_{\rm Band}$
324: for the Band function or
325: $\Delta \chi^{2}$ $\equiv \chi^{2}_{\rm PL} - \chi^{2}_{\rm CPL}$ for
326: a CPL model) greater than 6\footnote{This is a current criterion used in
327: the BAT team for reporting the spectral parameters based on a CPL fit in
328: the BAT refined circular of the Gamma-ray Burst Coordinates Network.} as
329: a function of the $\ep$ and the energy flux in the 15-150 keV band.
330: %%% 2nd ref comment
331: This $\Delta \chi^{2}$ $>$ 6 corresponds to $>$2.4 $\sigma$ confidence.
332: %%%
333: The figures in the left and right row show
334: the results based on the Band function and a CPL model, respectively.
335: We note the distinct differences in the shapes of the confidence contours,
336: especially
337: at low $\ep$, between the Band function and a CPL model as an input spectrum.
338: The results show that if a CPL model is indeed a true spectral shape,
339: BAT can measure $\ep$ at the lower boundary of its energy range
340: ($\sim$ 15 keV) with a very high significance. On the other hand, a low
341: $\ep$ measurement would be very challenging if the Band function is the
342: true spectral shape. Figure \ref{fig:exp_band_cpl_lowep} explains the
343: reason for these differences. The figure shows the calculated photon spectra
344: in a CPL model and the Band function for $\ep$ = 15 keV. In a CPL model,
345: the spectrum can not be fit with a PL model because of the curved shape
346: (exponential component) in the BAT observed energy band. Therefore, we
347: would expect a significant improvement in $\chi^{2}$ with a CPL
348: fit over a PL fit. However, in the Band function, due to the extra
349: power-law component (high energy power-law component) in the formula,
350: the spectrum at the BAT observed energy band would be just a simple
351: power-law with a high energy photon index. This is the reason why we
352: see a difference in the confidence contours based on the assumed
353: spectral models. The results also show that the $\ep$
354: measurement becomes difficult for BAT when $\ep$ is below 30 keV or
355: above 100 keV in the Band function shape. In the CPL shape, $\ep$ can be
356: determined even at $\sim$ 15 keV.
357:
358: %% 2nd ref comment
359: %Figure \ref{fig:sim_ep_fluence} also clearly shows that an energy fluence of
360: %$\sim$ $3 \times 10^{-6}$ ergs cm$^{2}$ is required to measure
361: %$\ep$ even when $\ep$ is located in the middle of the
362: %BAT observed energy band in the case of an on-axis burst. If a burst
363: %is detected at far off-axis such as 50$^{\circ}$, a fluence of
364: %$\sim$ $10^{-5}$ ergs cm$^{2}$ is needed to measure $\ep$ by the
365: %BAT. Another point which the simulations show is that the $\ep$
366: %measurement becomes difficult for BAT when $\ep$ is below 30 keV or
367: %above 100 keV in the Band function shape. In the CPL shape, $\ep$ can be
368: %determined even at $\sim$ 15 keV.
369: Figure \ref{fig:hist_pl_cpl} shows the number of the BAT GRBs which can be
370: acceptably fit by a PL model and by a CPL model as a function of the
371: 15-150 keV fluence. The data are from the BAT1 catalog.
372: In the case of an incident angle less than 25 degrees, a CPL
373: model becomes an acceptable fit for fluence $>$10$^{-6}$ ergs cm$^{-2}$.
374: However, a PL model still be acceptable fit if $\ep$ is located above or below
375: the BAT energy range. On the other hand, the fluence must typically be greater
376: than $3 \times 10^{-6}$ ergs cm$^{-2}$ in the case of an incident
377: angle greater than 50 degrees.
378: These threshold fluences required to measure $\ep$ in the BAT data
379: correspond to the $\sim$50\% confidence contour (green) in our
380: simulation results of Figure \ref{fig:sim_ep_fluence}.
381: %%%
382:
383: Next, we made histograms of $\ep$ for each photon index on a 0.1 grid
384: from 0 to 3.5 using the
385: %% 2nd ref comment
386: range of fluences corresponding to
387: the 1-$\sigma$ interval of the BAT observed fluence distribution
388: in the BAT1 catalog. The 1-$\sigma$ fluence
389: interval corresponds to the
390: range from $3.4 \times 10^{-7}$ ergs cm$^{-2}$ to $5.4 \times 10^{-6}$ ergs cm$^{-2}$.
391: This selection
392: of the fluence range allows us to reduce the systematic effect of the inclusion
393: of unrealistically bright or dim simulations.
394: %% 3rd ref comment
395: Furthermore, since we are interested in estimating the $\ep$ for the bursts which
396: do not show a significant improvement in $\chi^{2}$ by a CPL fit over a PL fit,
397: we also only selected the simulated spectra with $\Delta \chi^{2}$
398: $=$ $\chi^{2}_{\rm PL} - \chi^{2}_{\rm CPL}$ $< 6$. Because the numbers of simulated
399: spectra are different for each $\ep$ grid due to these selections, we normalized the
400: number of simulated spectra in each $\Gamma$-$\ep$ grid by the total number of spectra
401: in each $\ep$ grid.
402: %%
403: Figure \ref{fig:sim_phindex} shows the
404: contour map of the photon index ($\Gamma$) and $\log \ep$ for the Band function (left)
405: and a CPL (right) model. There is a
406: correlation between $\Gamma$ and $\log \ep$ in the range from 1.3 to 2.3 of
407: $\Gamma$ for the Band function. The correlation continues to $\Gamma = 3.0$ in the
408: case of a CPL model for the same reason as we demonstrated in Figure \ref{fig:exp_band_cpl_lowep}.
409: It might be interesting to note that a
410: very steep photon index such as $\Gamma$ $\sim$ 3 is not possible to achieve
411: if the source spectrum is the Band function. In this case, the source
412: spectrum might be much closer to a CPL shape. One important conclusion is that
413: the correlation between $\log \ep$ and $\Gamma$ exists independent of the
414: incident angle of the burst. Therefore, this correlation, the $\ep$ - $\Gamma$
415: relation, can be used for all BAT long GRBs within the allowed $\Gamma$ range,
416: %% 2nd ref comment
417: although with larger uncertainty for GRBs at large incidence angle.
418: %%
419:
420: We extracted the peak $\ep$ value from each histogram of $\Gamma$
421: and fit with a linear function using the range from $1.3 < \Gamma < 2.3$
422: for both the Band function and a CPL model.
423: Although the correlation exists until $\Gamma = 3$ in a CPL case,
424: we use the same $\Gamma$ range for the Band function and a CPL model
425: to investigate the systematic difference based on the assumption of the
426: source spectrum.
427: The best fit $\ep$-$\Gamma$ relations are summarized in Table
428: \ref{tbl:ep_gamma_band} (Band function) and Table \ref{tbl:ep_gamma_cpl} (CPL).
429: %
430: To estimate the 1-$\sigma$ uncertainty of the relation, we found 16\% and 84\%
431: points of $\ep$ from each histogram of $\Gamma$ and fitted with a cubic function
432: from $1.3 < \Gamma < 2.3$. The best fit cubic functions of the lower and higher
433: 1-$\sigma$ confidence level are also summarized in Table
434: \ref{tbl:ep_gamma_band} and Table \ref{tbl:ep_gamma_cpl}.
435: Figure \ref{fig:ep_gamma_fit} shows the best fit functions of
436: the $\ep$ - $\Gamma$ relation and its 1-$\sigma$ confidence level with the data points
437: used in the fittings.
438: %%% 2nd ref comment
439: We note that the wide $\ep$ range in the simulations (in our case from
440: 1.4 keV to 1210 keV) is essential to derive the 1-$\sigma$ confidence level
441: of the relation. If the $\ep$ range in the simulations is not wide enough
442: such as from 10 keV to 500 keV, we noticed that the confidence level will
443: be underestimated by a factor of 2 for the upper limit at $\Gamma$ of 1.3
444: and by a factor of 5 for the lower limit at $\Gamma$ of 2.0. Due to the
445: smoothly curved shapes of the Band function and the CPL model, the $\ep$ grids
446: in the simulations have to be an order of magnitude wider than the energy
447: range of the instrument, so that a curvature (or $\ep$)
448: in the spectrum is completely outside the energy range of the instrument
449: for the $\ep$ around the energy limits of the instrument.
450: However, we also notice that the best fit $\ep$ - $\Gamma$ relation itself
451: is less sensitive to the energy limits on the simulations.
452: %
453: Although the confidence level is different between
454: the Band function and a CPL model, the best fit linear function shows little
455: difference between these two spectral models.
456: We also calculated the relation weighting the
457: results at the incident angles of 0$^{\circ}$, 15$^{\circ}$,
458: 30$^{\circ}$, and 50$^{\circ}$ by the distribution of
459: the incident angle of the BAT GRBs (Figure \ref{fig:theta}). Hereafter,
460: we call this relation as the weighted $\ep$ - $\Gamma$ relation.
461: The contour plots of the weighted $\ep$ - $\Gamma$ relation, the plot of the
462: best fit functions, and the formula of the best fit functions are shown
463: and summarized in Figure \ref{fig:weighted_ep_gamma}, Figure
464: \ref{fig:weighted_ep_gamma_fit}, and Table \ref{tbl:ep_gamma_band} and
465: Table \ref{tbl:ep_gamma_cpl}, respectively.
466:
467: %%% 2nd ref comment
468: In the application of our $\ep$ - $\Gamma$ relation, we strongly encourage
469: the reader to use the result based on the Band function as a prior. The main
470: reason for also performing the simulations of a CPL model as a prior is to see the
471: systematic effect due to a prior assumption of the spectral model. Our results
472: are clearly demonstrating the effect of the assumed spectral model. From the
473: various measurements of the burst spectra by different instruments, the true
474: burst spectrum is very likely to be the Band function at least for long GRBs.
475: Therefore, the $\ep$ - $\Gamma$ relation based on the Band function as a prior
476: is the most suitable relation to apply for the BAT long GRBs.
477:
478: \section{Comparison to other $\ep$ Measurements}
479:
480: To investigate the validity of our simulation study, we used the
481: spectral parameters on the BAT1 catalog \citep{sakamoto2007}.
482: Table \ref{hyo:spec_para} shows the spectral parameters of 31 long GRBs (T$_{90}$ $>$
483: 2 seconds) having $\Delta \chi^{2}$ of greater than 6 in a CPL model over a PL model fit.
484: Figure \ref{fig:comp_epeak_gamma_obs} shows $\ep$ energy in a CPL model
485: and the photon index, $\Gamma$, in a PL model for the BAT GRBs overlaid with the
486: weighted $\ep$ - $\Gamma$ relation. We also plot $\Gamma$ derived
487: from the BAT data and $\ep$ reported by Konus-Wind or $HETE$-2 in the Gamma ray
488: bursts Coordinates Network (GCN) listed in table \ref{hyo:spec_para_kw_hete2}.
489: As seen in the figure,
490: %% 2nd ref comment
491: the 1-$\sigma$ confidence level of the $\ep$ - $\Gamma$ relation based on the simulation study
492: is consistent with the 90\% confidence level of $\Gamma$ observed by the BAT
493: and $\ep$ observed by the GRB instruments.
494: %%
495: However, we want to caution about using our $\ep$ - $\Gamma$
496: relation for estimating $\ep$. Our estimator is based on prior
497: assumptions of the low-energy and/or the high-energy photon index
498: measured by the BATSE. Therefore, $\ep$ based on our $\ep$ - $\Gamma$
499: relation only provides a likelihood of the $\ep$ value not the ``measurement.''
500:
501: %% 2nd ref comment
502: The calculation of the bolometric flux or fluence is another challenge
503: when using the
504: BAT data alone. However, since the low energy photon index $\alpha$ and the high energy
505: photon index $\beta$ of the Band function are quite stable parameters even if $\ep$ varies from
506: a few keV to a few MeV \citep[e.g.,][]{sakamoto2005,kaneko2006}, one could
507: estimate the bolometric flux or fluence assuming the best fit $\alpha$ and $\beta$
508: from the BATSE time-averaged spectral analysis of \citet{kaneko2006}, and applying
509: the best fit $\ep$ derived from our $\ep$ - $\Gamma$ relation.
510: To get the normalization for the Band function spectrum, one would scale the Band
511: spectrum so that the flux in the BAT energy range matches to the BAT measured flux.
512: One can also estimate
513: the error of the flux or fluence by propagating the errors of $\alpha$, $\beta$, $\ep$,
514: and the normalization, however this estimate will not be strictly correct because
515: the parameters of the Band function are correlated. Finally, we caution against relying
516: too heavily on this derived bolometric flux or fluence, since the method described uses
517: averaged $\alpha$ and $\beta$ from a different burst population and an estimated,
518: rather than measured $\ep$.
519:
520: \section{Discussion}
521:
522: According to \citet{sakamoto2005}, an equal number of
523: X-ray flashes (XRF), X-ray-rich GRBs (XRR), and GRBs are
524: reported in the HETE-2 GRB sample. Their classification of GRBs based on the fluence
525: ratio between the 2-30 keV and 30-400 keV bands is almost the equivalent of classifying
526: GRBs by $\ep$. The boundaries of $\ep$ between an XRF and
527: an XRR, and an XRR and a GRB are around 30 keV and 100 keV.
528: When we use the weighted $\ep$ - $\Gamma$ relation for the Band function to calculate
529: the corresponding $\Gamma$ for each $\ep$, $\Gamma$ is $\sim$ 2.2 and
530: $\sim$ 1.5 for $\ep$ of 30 keV and 100 keV, respectively. Applying these $\Gamma$
531: criteria to a sample of 206 BAT
532: GRBs, excluding short GRBs (T$_{90}$ $<$ 2 seconds) and GRBs with incomplete dataset,
533: we found that the number of XRFs, XRRs, and GRBs are 20, 126, and 60 respectively.
534: The numbers of XRFs, XRRs, and GRBs in the HETE-2 sample are 16, 19, and 10
535: respectively. Therefore, the ratio of the numbers of XRRs and GRBs is identical
536: for both BAT and HETE-2 sample.
537: %Thus, the numbers of
538: %BAT GRBs classified as XRR and GRB are almost identical.
539: The small numbers of XRFs in the BAT sample is due to the difficulty in observing
540: very soft XRFs in the BAT \citep{band2003,band2006}.
541: However, as mentioned in \citet{band2006}, it is
542: very difficult to determine the actual detection threshold of the BAT due to its
543: complexity in the triggering algorithm. Although nothing could be addressed about
544: the actual number of XRFs, the number of GRBs in XRRs and GRBs seen in
545: the BAT sample is consistent with the HETE-2 sample. The detailed study of the
546: Swift XRFs and XRRs is presented elsewhere \citep{swift_xrfxrr}.
547:
548: \citet{butler2007} calculated $\ep$ by their Bayesian approach for 218 {\it Swift}
549: GRBs using only the BAT data. Based on their calculated $\ep$ and the bolometric fluence,
550: they claimed that all of the empirical relations, $\eps$ - $\eiso$ \citep{amati2002},
551: $\eps$ - $\lpiso$ \citep{yonetoku2004}, and $\eps$$T_{45}$ - $\lpiso$ \citep{firmani2006},
552: proposed in the pre-{\it Swift} observations are not valid for the {\it Swift} BAT sample.
553: We investigated the validity of their $\ep$ by checking $\ep$ obtained by using our
554: $\ep$ - $\Gamma$ relation.
555: %
556: We created the BAT spectra for their GRB samples by the time interval
557: reported on Table 1 of \citet{butler2007}. Then, we fit the spectrum by a PL
558: model to extract the best fit $\Gamma$. By only selecting their $\Gamma$ within
559: the allowed $\Gamma$ range for applying our $\ep$ - $\Gamma$ relation ($1.3 < \Gamma
560: < 2.3$) and also excluding the short GRBs (156 samples in total),
561: we calculated $\ep$ applying our weighted $\ep$ - $\Gamma$ relation for
562: the Band function.
563: %
564: %
565: Figure \ref{fig:comp_taka_nat_ep} shows the $\ep$ reported on
566: \citet{butler2007} versus $\ep$ derived from our weighted $\ep$ - $\Gamma$ relation for
567: the Band function.
568: %% 2nd ref comment
569: Although the error bars are large in both estimators,
570: the figure shows that $\ep$ of the \citet{butler2007} sample
571: has a systematically higher $\ep$ compared to that from our $\ep$ - $\Gamma$ relation.
572: %
573: About 20\% of the \citet{butler2007} sample selected based on the
574: range of $\Gamma$ from 1.3 to 2.3 exceeds $\ep$ $\sim$ 150 keV which is the limit
575: of the estimated $\ep$ using our $\ep$ - $\Gamma$ relation for $\Gamma$ = 1.3.
576: %% 2nd ref comment
577: Furthermore, we are already excluding 20\% of the \citet{butler2007} sample because
578: those bursts fall
579: outside limit range of $\Gamma$ from 1.3 to 2.3 in our relation. This
580: limit is determined because $\ep$ is very likely located outside of the BAT energy
581: range, and therefore, the BAT data alone can not constrain about $\ep$ (the BAT
582: data only can provide the limit in $\ep$). In total, about 35\% of the \citet{butler2007}
583: samples are obviously inconsistent with the $\ep$ estimated based on our
584: $\ep$ - $\Gamma$ relation.
585: However, $\ep$ is {\it constrained} in the most of
586: the \citet{butler2007} sample. These results provide a caution for the method for
587: estimating $\ep$ in \citet{butler2007}.
588: %
589: % also plot the BAT GRB sample for which the $\ep$ can be determined from a CPL model
590: %from the BAT data alone (Table \ref{hyo:spec_para}). Interestingly,
591: %$\ep$ sample which can be measured by the BAT data alone agrees quite nicely to
592: %$\ep$ of \citet{butler2007}. This suggests that even in the
593: %\citet{butler2007} approach there might be a systematic problem
594: %in calculating $\ep$ if the BAT spectral data itself does not require any
595: %spectral model other than a PL model.
596:
597: %% 2nd ref comment
598: \citet{butler2007} justify their $Swift$-only $\ep$ estimates in part by
599: comparing to {\it Konus-Wind} measurements of $\ep$ for the same bursts. They also
600: using the $\ep$ distribution as a prior.
601: %
602: However, the assumption of $\ep$ measured by {\it Konus-Wind} should be identical
603: to that of BAT in \citet{butler2007} might not be valid. Because BAT has a significantly
604: larger effective area and also relatively softer energy band than {\it Konus-Wind}, the
605: time interval for creating the time-averaged spectrum based on the BAT data could be
606: systematically longer than that of {\it Konus-Wind} (Sakamoto et al. in preparation).
607: This longer time interval for the time-averaged spectrum in BAT might lead to a
608: systematically lower $\ep$ which might contradict with the $\ep$ based on the {\it Konus-Wind}
609: data alone.
610: %% 2nd ref comment
611: %This systematic bias of $\ep$ measured by different instruments is already
612: %seen in Figure 10. The $\ep$ measured by the BAT data alone is systematically lower than
613: %the $\ep$ measured by {\it Konus-Wind} or {\it HETE-2}.
614: %% 3rd ref comment
615: For instance, GRB 060117, which has individual measurements of $\ep$ from the BAT
616: and the {\it Konus-Wind} data (see Table 2 and Table 3), shows a smaller
617: $\ep$ in the BAT data. The duration reported based on the {\it Konus-Wind} data is
618: $\sim$ 20 seconds \citep{golenetskii2006b}. On the other hand, the duration used to accumulate
619: the BAT spectrum is $\sim$ 30 seconds \citep{sakamoto2007}. We confirmed based on
620: our cross-calibration work that there is no systematic difference in $\ep$ of this burst
621: between BAT and {\it Konus-Wind} if we select exactly the same time interval for accumulating
622: the spectrum (Sakamoto et al. in preparation). Therefore, we believe that a prior assumption
623: of $\ep$ based on a particular GRB instrument might introduce an another level of a systematic
624: error in the analysis.
625: %
626: Most importantly, we believe that testing these
627: empirical relations, which require the broad-band spectral properties of the prompt GRB
628: emission, by using only the BAT narrow-band data could lead to a wrong conclusion. Current on-going
629: activity for analyzing the spectral data of simultaneously observed BAT GRBs by other
630: GRB missions such as {\it Konus-Wind} and {\it Suzaku}/WAM (Sakamoto et al. in
631: preparation; Krimm et al. in preparation) is indeed a necessary step to answer for
632: the validation of these empirical relations.
633: %% 2nd ref comment
634: We might want to emphasize that the 1-$\sigma$ confidence level of our $\ep-\Gamma$
635: relation based on the Band function includes most of $\ep$ reported by other
636: instruments (see Figure \ref{fig:comp_epeak_gamma_obs}). Therefore, the confidence level which we
637: are quoting in our estimator is large enough to include the systematic problem
638: in $\ep$ among the different instruments.
639: %
640:
641: We report the correlation between $\ep$ and the photon index, $\Gamma$,
642: of the BAT prompt emission spectrum based on our simulation study.
643: Using this relation, it is possible to estimate $\ep$ from $\Gamma$ in the
644: range from 1.3 to 2.3.
645: %%
646: We also performed the spectral simulations for assuming various incident
647: angles (0$^{\circ}$, 15$^{\circ}$, 30$^{\circ}$ and 50$^{\circ}$) and
648: different spectral models (Band function and CPL). However, none of these
649: systematic effects changes the relation.
650: %%
651: %% 2ne ref comment
652: In the application, the $\ep$ - $\Gamma$ relation based on the Band function
653: as a prior is the appropriate formula to use.
654: %
655: The $\ep$ - $\Gamma$ relation could be informative
656: for classifying
657: the BAT GRBs from the photon index alone as derived from a simple power-law
658: model which is the best fit for about 80 \% of the whole population of the
659: BAT GRBs.
660:
661: \acknowledgements
662: We would like to thank the anonymous referee for comments and suggestions
663: that materially improved the paper.
664:
665: \begin{thebibliography}{}
666: \bibitem[Amati et al.(2002)]{amati2002} Amati, L., et al. 2002, A\&A,
667: 390, 81
668: \bibitem[Amati(2003)]{amati2003} Amati, L., 2003, ChJAA, Vol. 3, Supplement, pp.
669: 455-460
670: \bibitem[Barthelmy et al.(2005)]{barthelmy2005} Barthelmy, S.D., et al. 2005,
671: Space Sci. Rev., 120, 143
672: \bibitem[Band et al.(1993)]{band1993}
673: Band, D. L., et al. 1993, \apj, 413, 281
674: \bibitem[Band (2003)]{band2003}
675: Band, D. L., 2003, ApJ, 588, 945
676: \bibitem[Band (2006)]{band2006}
677: Band, D. L., 2006, ApJ, 664, 378
678: \bibitem[Barraud, et al.(2005)]{barraud2005}
679: Barraud, C., Daigne, F., Mochkovitch, R., Atteia, J. L. 2005,
680: A\&A, 440, 809
681: \bibitem[Butler et al.(2007)]{butler2007}
682: Butler, N.R., Kocevski, D., Bloom, J.S., Curtis, J.L. 2007, \apj, 671, 656
683: \bibitem[Crew et al.(2005a)]{crew2005a} Crew, G., et al. 2005a, GCN Circ. 3890, http://gcn.gsfc.nasa.gov/gcn3/3890.gcn3
684: \bibitem[Crew et al.(2005b)]{crew2005b} Crew, G., et al. 2005b, GCN Circ. 4021, http://gcn.gsfc.nasa.gov/gcn3/4021.gcn3
685: \bibitem[Dermer et al.(1999)]{dermer1999}
686: Dermer, C. D., Chiang, J., and B$\ddot{\rm o}$ttcher
687: 1999, \apj, 513, 656
688: \bibitem[Dermer and Mitman(2003)]{dermer2003}
689: Dermer, C. D., and Mitman, K. E. 2003, in ASP Conf. Ser. 312, Third
690: Rome Workshop on Gamma-Ray Bursts in the Afterglow Era, ed. M. Feroci
691: et al. (San Francisco: ASP), 301
692: \bibitem[Firmani et al.(2006)]{firmani2006}
693: Firmani, C, Ghisellini, G., Avila-Reese, V., and Ghirlanda, G.
694: 2006, MNRAS, 370, 185
695: \bibitem[Gehrels et al (2004)]{gehrels2004} Gehrels, N., et al. 2004,
696: ApJ, 611, 1005
697:
698: \bibitem[Ghirlanda et al.(2004)]{ghirlanda2004} Ghirlanda, G.,
699: Ghisellini, G., Lazzati, D., 2004, ApJ, 616, 331
700: \bibitem[Golenetskii et al.(2005a)]{golenetskii2005a} Golenetskii, S., et al. 2005a, GCN Circ. 3152, http://gcn.gsfc.nasa.gov/gcn3/3152.gcn3
701: \bibitem[Golenetskii et al.(2005b)]{golenetskii2005b} Golenetskii, S., et al. 2005b, GCN Circ. 3474, http://gcn.gsfc.nasa.gov/gcn3/3474.gcn3
702: \bibitem[Golenetskii et al.(2005c)]{golenetskii2005c} Golenetskii, S., et al. 2005c, GCN Circ. 3518, http://gcn.gsfc.nasa.gov/gcn3/3518.gcn3
703: \bibitem[Golenetskii et al.(2005d)]{golenetskii2005d} Golenetskii, S., et al. 2005d, GCN Circ. 3619, http://gcn.gsfc.nasa.gov/gcn3/3619.gcn3
704: \bibitem[Golenetskii et al.(2005e)]{golenetskii2005e} Golenetskii, S., et al. 2005e, GCN Circ. 4078, http://gcn.gsfc.nasa.gov/gcn3/4078.gcn3
705: \bibitem[Golenetskii et al.(2005f)]{golenetskii2005f} Golenetskii, S., et al. 2005f, GCN Circ. 4238, http://gcn.gsfc.nasa.gov/gcn3/4238.gcn3
706: \bibitem[Golenetskii et al.(2006a)]{golenetskii2006a} Golenetskii, S., et al. 2006a, GCN Circ. 4439, http://gcn.gsfc.nasa.gov/gcn3/4439.gcn3
707: \bibitem[Golenetskii et al.(2006b)]{golenetskii2006b} Golenetskii, S., et al. 2006b, GCN Circ. 4542, http://gcn.gsfc.nasa.gov/gcn3/4542.gcn3
708: \bibitem[Golenetskii et al.(2006c)]{golenetskii2006c} Golenetskii, S., et al. 2006c, GCN Circ. 4881, http://gcn.gsfc.nasa.gov/gcn3/4881.gcn3
709: \bibitem[Golenetskii et al.(2006d)]{golenetskii2006d} Golenetskii, S., et al. 2006d, GCN Circ. 5113, http://gcn.gsfc.nasa.gov/gcn3/5113.gcn3
710: \bibitem[Golenetskii et al.(2006e)]{golenetskii2006e} Golenetskii, S., et al. 2006e, GCN Circ. 5446, http://gcn.gsfc.nasa.gov/gcn3/5446.gcn3
711: \bibitem[Golenetskii et al.(2006f)]{golenetskii2006f} Golenetskii, S., et al. 2006f, GCN Circ. 5460, http://gcn.gsfc.nasa.gov/gcn3/5460.gcn3
712: \bibitem[Golenetskii et al.(2006g)]{golenetskii2006g} Golenetskii, S., et al. 2006g, GCN Circ. 5518, http://gcn.gsfc.nasa.gov/gcn3/5518.gcn3
713: \bibitem[Golenetskii et al.(2006h)]{golenetskii2006h} Golenetskii, S., et al. 2006h, GCN Circ. 5722, http://gcn.gsfc.nasa.gov/gcn3/5722.gcn3
714: \bibitem[Golenetskii et al.(2006i)]{golenetskii2006i} Golenetskii, S., et al. 2006i, GCN Circ. 5748, http://gcn.gsfc.nasa.gov/gcn3/5748.gcn3
715: \bibitem[Golenetskii et al.(2006j)]{golenetskii2006j} Golenetskii, S., et al. 2006j, GCN Circ. 5837, http://gcn.gsfc.nasa.gov/gcn3/5837.gcn3
716: \bibitem[Golenetskii et al.(2006k)]{golenetskii2006k} Golenetskii, S., et al. 2006k, GCN Circ. 5890, http://gcn.gsfc.nasa.gov/gcn3/5890.gcn3
717: \bibitem[Golenetskii et al.(2006l)]{golenetskii2006l} Golenetskii, S., et al. 2006l, GCN Circ. 5984, http://gcn.gsfc.nasa.gov/gcn3/5984.gcn3
718: \bibitem[Golenetskii et al.(2007a)]{golenetskii2007a} Golenetskii, S., et al. 2007a, GCN Circ. 6124, http://gcn.gsfc.nasa.gov/gcn3/6124.gcn3
719: \bibitem[Golenetskii et al.(2007b)]{golenetskii2007b} Golenetskii, S., et al. 2007b, GCN Circ. 6230, http://gcn.gsfc.nasa.gov/gcn3/6230.gcn3
720: \bibitem[Golenetskii et al.(2007c)]{golenetskii2007c} Golenetskii, S., et al. 2007c, GCN Circ. 6344, http://gcn.gsfc.nasa.gov/gcn3/6344.gcn3
721: \bibitem[Golenetskii et al.(2007d)]{golenetskii2007d} Golenetskii, S., et al. 2007d, GCN Circ. 6403, http://gcn.gsfc.nasa.gov/gcn3/6403.gcn3
722: \bibitem[Golenetskii et al.(2007e)]{golenetskii2007e} Golenetskii, S., et al. 2007e, GCN Circ. 6459, http://gcn.gsfc.nasa.gov/gcn3/6459.gcn3
723: \bibitem[Huang et al.(2002)]{huang2002}
724: Huang, Y. F., Dai, Z. G., and Lu, T. 2002,
725: MNRAS, 332, 735
726: \bibitem[Kaneko et al.(2006)]{kaneko2006}
727: Kaneko, Y. et al. 2006, ApJS, 166, 298
728: \bibitem[Kippen et al.(2002)]{kippen_xrf_astroph}
729: Kippen, R. M., Woods, P. M., Heise, J., in't Zand, J., Briggs,
730: M. S., \& Preece, R. D. 2002, in Gamma-Ray Bursts and
731: Afterglow Astronomy, eds. G. R. Ricker and R. Vanderspek (New
732: York: AIP), 244
733: \bibitem[Lamb, Donaghy \& Graziani(2005)]{lamb2005}
734: Lamb, D. Q., Donaghy, T. Q., and Graziani, C. 2005, ApJ, 520,
735: 335
736: \bibitem[Liang \& Zhang(2005)]{liang2005} Liang, E., Zhang, B., ApJ,
737: 633, 611
738: \bibitem[M\'esz\'aros et al.(2002)]{mes2002} M\'esz\'aros, P., Ramirez-Ruiz, E.,
739: Rees, M. J., Zhang, B., ApJ, 578, 812
740: \bibitem[Mochkovitch et al.(2003)]{mochkovitch2003}
741: Mochkovitch, R., Daigne, F., Barraud, C., \& Atteia, J. L.
742: 2003, in APS Conf. Ser. 312, Third Rome Workshop on Gamma-Ray Bursts in
743: the Afterglow Era, ed. M. Feroci et al. (San Francisco: ASP), 381
744: \bibitem[Nakagawa et al.(2005)]{nakagawa2005} Nakagawa, U., et al. 2005, GCN Circ. 3053, http://gcn.gsfc.nasa.gov/gcn3/3053.gcn3
745: \bibitem[Rossi et al.(2002)]{rossi2002}
746: Rossi, E., Lazzati, D., and Rees, M. J. 2002, MNRAS,
747: 332, 945
748: \bibitem[Sakamoto et al.(2004)]{sakamoto2004} Sakamoto, T., et al. 2004, ApJ, 602, 875
749: \bibitem[Sakamoto et al.(2005)]{sakamoto2005} Sakamoto, T., et al. 2005,
750: ApJ, 629, 311
751: \bibitem[Sakamoto et al.(2006)]{sakamoto2006} Sakamoto, T., et al. 2006, ApJ, 636, L73
752: \bibitem[Sakamoto et al.(2008a)]{sakamoto2007} Sakamoto, T., et
753: al. 2008a, ApJS, 175, 179
754: \bibitem[Sakamoto et al.(2008b)]{swift_xrfxrr} Sakamoto, T., et al. 2008b, ApJ, 679, 570
755: \bibitem[Sato et al.(2005)]{rsato2005} Sato, R., et al. 2005, PASJ, 57, 1031
756: \bibitem[Toma et al.(2005)]{toma2005} Toma, K., Yamazaki, R., Nakamura,
757: T. 2005, ApJ, 635, 481
758: \bibitem[Yamazaki et al.(2004)]{yamazaki2004} Yamazaki, R., Ioka, K.,
759: Nakamura, T. 2004, ApJ, 607, L103
760: \bibitem[Yonetoku et al.(2004)]{yonetoku2004} Yonetoku, D., et al. 2004, ApJ, 609, 935
761: \bibitem[Zhang \& M\'esz\'aros(2002)]{zhang2002}
762: Zhang, B. \& M\'esz\'aros, P. 2002, \apj, 571, 876
763: \bibitem[Zhang et al.(2004)]{zhang2004}
764: Zhang, B., Dai, X., Lloyd-Ronning, N. M., \& M\'esz\'aros, P.
765: 2004, 601, L119
766:
767: \end{thebibliography}
768:
769: \clearpage
770:
771: %% tables
772: %%% Ep-Gamma relations - Band function -
773: \begin{deluxetable}{ccccc}
774: \tabletypesize{\footnotesize}
775: \rotate
776: \tablecaption{$\ep$ - $\Gamma$ relation based on the Band function\label{tbl:ep_gamma_band}}
777: \tablewidth{0pt}
778: \tablehead{
779: \colhead{$\theta$} &
780: \colhead{} &
781: \colhead{$\ep-\Gamma$ relation} &
782: \colhead{1-$\sigma$ lower limit} &
783: \colhead{1-$\sigma$ upper limit}
784: }
785: \startdata
786: 0 (on-axis)& $\log \ep =$ & $3.312 - 0.817\Gamma$ & $-29.450 + 57.904\Gamma -34.337\Gamma^{2} + 6.445\Gamma^{3}$ &
787: $-1.073 + 9.840\Gamma -7.065\Gamma^{2} + 1.413\Gamma^{3}$\\
788: 15 & $\log \ep =$ & $3.184 - 0.793\Gamma$ & $-31.986 + 62.511\Gamma -37.070\Gamma^{2} + 6.975\Gamma^{3}$ &
789: $-1.991 + 11.452\Gamma -7.988\Gamma^{2} + 1.587\Gamma^{3}$\\
790: 30 & $\log \ep =$ & $3.231 - 0.819\Gamma$ & $-20.684 + 43.646\Gamma -26.891\Gamma^{2} + 5.185\Gamma^{3}$ &
791: $-6.762 + 19.192\Gamma -12.065\Gamma^{2} + 2.291\Gamma^{3}$\\
792: 50 & $\log \ep =$ & $3.210 - 0.796\Gamma$ & $6.782 - 3.948\Gamma -0.286 \Gamma^{2} + 0.348\Gamma^{3}$ &
793: $-6.860 + 18.110\Gamma -10.740\Gamma^{2} + 1.935\Gamma^{3}$\\
794: Weighted & $\log \ep =$ & $3.258 -0.829 \Gamma$ & $-20.684 + 43.646\Gamma -26.891\Gamma^{2} + 5.185\Gamma^{3}$ &
795: $-5.198 + 16.568\Gamma -10.630\Gamma^{2} + 2.034\Gamma^{3}$\\
796: \enddata
797: \end{deluxetable}
798:
799: %% tables
800: %%% Ep-Gamma relations - CPL model -
801: \begin{deluxetable}{ccccc}
802: \tabletypesize{\footnotesize}
803: \rotate
804: \tablecaption{$\ep$ - $\Gamma$ relation based on a CPL model\label{tbl:ep_gamma_cpl}}
805: \tablewidth{0pt}
806: \tablehead{
807: \colhead{$\theta$} &
808: \colhead{} &
809: \colhead{$\ep-\Gamma$ relation} &
810: \colhead{1-$\sigma$ lower limit} &
811: \colhead{1-$\sigma$ upper limit}
812: }
813: \startdata
814: 0 (on-axis)& $\log \ep =$ & $3.722 - 1.033\Gamma$ & $1.829 +1.874\Gamma - 1.638\Gamma^{2} + 0.315\Gamma^{3}$ &
815: $-14.504 + 31.357\Gamma -18.043\Gamma^{2} + 3.234\Gamma^{3}$\\
816: 15 & $\log \ep =$ & $3.657 - 0.994\Gamma$ & $1.829 + 1.874\Gamma - 1.638\Gamma^{2} +0.315\Gamma^{3}$ &
817: $-14.504 + 31.357\Gamma -18.043\Gamma^{2} + 3.234\Gamma^{3}$\\
818: 30 & $\log \ep =$ & $3.490 - 0.904\Gamma$ & $1.980 + 1.434\Gamma -1.342\Gamma^{2} + 0.258\Gamma^{3}$ &
819: $-8.724 + 20.829\Gamma -11.819\Gamma^{2} + 2.052\Gamma^{3}$\\
820: 50 & $\log \ep =$ & $3.664 - 0.984\Gamma$ & $-0.742 + 5.847\Gamma -3.751 \Gamma^{2} + 0.695\Gamma^{3}$ &
821: $-1.794 + 8.489\Gamma -4.823\Gamma^{2} + 0.813\Gamma^{3}$\\
822: Weighted & $\log \ep =$ & $3.518 -0.920 \Gamma$ & $5.018 -3.548\Gamma +1.366\Gamma^{2} - 0.229\Gamma^{3}$ &
823: $-9.443 + 22.037\Gamma -12.478\Gamma^{2} + 2.168\Gamma^{3}$\\
824: \enddata
825: \end{deluxetable}
826:
827: %%% BAT spectral parameters
828: \begin{table}
829: \caption{The BAT time-averaged spectral parameters fitted with a simple power-law
830: (PL) model and a power-law times exponential cutoff (CPL) model. See the BAT1
831: catalog paper for the details about the BAT analysis (Sakamoto et
832: al. 2008a). The degree of freedom in a PL fit and a CPL fit is all 57
833: and 56 respectively.}
834: \vspace{0.5cm}
835: \centerline{
836: \label{hyo:spec_para}
837: {\small
838: \begin{tabular}{cccc|ccc}\\\hline\hline
839: & & \multicolumn{2}{c}{PL} & \multicolumn{3}{c}{CPL}\\\cline{3-7}
840: GRB & Trigger ID & $\Gamma$ & $\chi^{2}$ & $\alpha$ & $\ep$ [keV] &
841: $\chi^{2}$ \\\hline
842: % new
843: GRB 041217 & 100116 & $1.46 \pm 0.07$ & 74.9 & $-0.7 \pm 0.3$ & $95_{-14}^{+27}$ & 54.8\\
844: GRB 041224 & 100703 & $1.72 \pm 0.06$ & 56.1 & $-1.1 \pm 0.3$ & $74_{-9}^{+16}$ & 36.7\\
845: GRB 050117 & 102861 & $1.50 \pm 0.04$ & 38.8 & $-1.2 \pm 0.2$ & $143_{-33}^{+108}$ & 29.6\\
846: GRB 050124 & 103647 & $1.47 \pm 0.08$ & 58.7 & $-0.7 \pm 0.4$ & $95_{-16}^{+39}$ & 45.4\\
847: GRB 050128 & 103906 & $1.37 \pm 0.07$ & 59.3 & $-0.7 \pm 0.3$ & $113_{-19}^{+46}$ & 44.8\\
848: GRB 050219A & 106415 & $1.31 \pm 0.06$ & 103.2 & $-0.1 \pm 0.3$ & $92_{-8}^{+12}$ & 45.5\\
849: GRB 050219B & 106442 & $1.53 \pm 0.05$ & 86.6 & $-1.0_{-0.2}^{+0.3}$ & $108_{-16}^{+35}$ & 69.0\\
850: GRB 050410 & 114299 & $1.65 \pm 0.08$ & 78.5 & $-0.8 \pm 0.4$ & $74_{-9}^{+19}$ & 61.3\\
851: GRB 050416B & 114797 & $1.4 \pm 0.1$ & 67.4 & $-0.4_{-0.6}^{+0.7}$ & $94_{-19}^{+66}$ & 59.7\\
852: GRB 050525A$^{a}$ & 130088 & 1.76 & 166.4 & $-1.0 \pm 0.1$ & $82_{-3}^{+4}$ & 17.9\\
853: GRB 050716 & 146227 & $1.37 \pm 0.06$ & 52.5 & $-0.8 \pm 0.3$ & $123_{-24}^{+61}$ & 39.4\\
854: GRB 050815 & 150532 & $1.8 \pm 0.2$ & 75.6 & $0.9_{-1.4}^{+1.9}$ & $44_{-6}^{+9}$ & 62.1\\
855: GRB 050820B & 151334 & $1.34 \pm 0.04$ & 89.6 & $-0.6 \pm 0.2$ & $111_{-13}^{+21}$ & 48.7\\
856: GRB 050915B & 155284 & $1.90 \pm 0.06$ & 55.5 & $-1.4 \pm 0.3$ & $61_{-8}^{+17}$ & 46.0\\
857: GRB 051021B & 160672 & $1.6 \pm 0.1$ & 56.9 & $-0.6_{-0.6}^{+0.8}$ & $72_{-13}^{+45}$ & 49.7\\
858: GRB 060111A & 176818 & $1.65 \pm 0.07$ & 69.0 & $-0.9 \pm 0.3$ & $74_{-10}^{+19}$ & 50.4\\
859: GRB 060115 & 177408 & $1.8 \pm 0.1$ & 52.6 & $-1.0_{-0.5}^{+0.6}$ & $63_{-11}^{+36}$ & 45.8\\
860: GRB 060117 & 177666 & $1.93 \pm 0.03$ & 67.0 & $-1.5 \pm 0.1$ & $70_{-5}^{+7}$ & 35.6\\
861: GRB 060204B & 180241 & $1.44 \pm 0.09$ & 47.0 & $-0.8 \pm 0.4$ & $100_{-21}^{+75}$ & 38.9\\
862: GRB 060206 & 180455 & $1.71 \pm 0.08$ & 64.6 & $-1.2 \pm 0.3$ & $78_{-13}^{+38}$ & 55.3\\
863: GRB 060211A & 181126 & $1.8 \pm 0.1$ & 71.5 & $-0.9_{-0.5}^{+0.6}$ & $58_{-8}^{+18}$ & 60.6\\
864: GRB 060322 & 202442 & $1.58 \pm 0.07$ & 64.6 & $-1.1_{-0.4}^{+0.3}$ & $96_{-18}^{+90}$ & 57.5\\
865: GRB 060428B & 207399 & $2.6 \pm 0.2$ & 66.7 & $-0.8_{-1.2}^{+1.6}$ & $22_{-13}^{+5}$ & 59.1\\
866: GRB 060707 & 217704 & $1.7 \pm 0.1$ & 70.5 & $-0.6_{-0.6}^{+0.7}$ & $63_{-10}^{+21}$ & 60.5\\
867: GRB 060813 & 224364 & $1.36 \pm 0.04$ & 54.1 & $-1.0 \pm 0.2$ & $168_{-39}^{+117}$ & 43.5\\
868: GRB 060825 & 226382 & $1.72 \pm 0.07$ & 64.7 & $-1.2 \pm 0.3$ & $73_{-11}^{+28}$ & 53.7\\
869: GRB 060908 & 228581 & $1.35 \pm 0.06$ & 50.7 & $-1.0 \pm 0.3$ & $151_{-41}^{+184}$ & 44.2\\
870: GRB 060927 & 231362 & $1.65 \pm 0.08$ & 70.4 & $-0.9 \pm 0.4$ & $72_{-11}^{+25}$ & 57.5\\
871: GRB 070420 & 276321 & $1.56 \pm 0.05$ & 60.7 & $-1.2 \pm 0.2$ & $120_{-24}^{+76}$ & 51.1\\
872: GRB 070508 & 278854 & $1.35 \pm 0.03$ & 38.4 & $-1.1 \pm 0.1$ & $260_{-68}^{+203}$ & 27.8\\
873: GRB 070521 & 279935 & $1.36 \pm 0.04$ & 57.5 & $-1.1 \pm 0.2$ & $209_{-60}^{+234}$ & 50.1\\\hline
874: \end{tabular}}
875: }
876: \tablenotetext{a}{The confidence interval is not calculated because of $\chi^{2}_{\nu}$ $>$ 2.}
877: \tablenotetext{}{Short GRBs, GRB 050820A and GRB 050925 are excluded.}
878: \end{table}
879:
880: %% spectral parameters reported by K-W and HETE-2
881: \begin{table}
882: \caption{The spectral parameters of simultaneously observed by {\it Konus-Wind} or
883: {\it HETE-2}.}
884: \vspace{0.5cm}
885: \label{hyo:spec_para_kw_hete2}
886: \centerline{
887: {\small
888: \begin{tabular}{ccccccc}\\\hline\hline
889: GRB & Model & $\alpha$ & $\beta$ & $\ep$ & $\Gamma_{\rm{BAT}}$ & Reference\\\hline
890: GRB 050215B & Band & - & - & $<$ 35.7$^{a}$ & $2.0 \pm 0.2$ & Nakagawa et al. (2005)\\
891: GRB 050326 & Band & $-0.74 \pm 0.09$ & $-2.49 \pm 0.16$ & $201 \pm 24$ & $1.25 \pm 0.04$ & Golenetskii et al. (2005a)\\
892: GRB 050525A & CPL & $-1.10 \pm 0.05$ & - & $84.1 \pm 1.7$ & 1.76$^{c}$ & Golenetskii et al. (2005b)\\
893: GRB 050603 & Band & $-0.79 \pm 0.06$ & $-2.15 \pm 0.09$ & $349 \pm 28$ & $1.16 \pm 0.06$ & Golenetskii et al. (2005c)\\
894: GRB 050713A & CPL & $-1.12 \pm 0.08$ & - & $312 \pm 50$ & $1.53 \pm 0.08$ & Golenetskii et al. (2005d)\\
895: GRB 050824 & Band & - & - & $<$ 12.7$^{b}$ & $2.8 \pm 0.4$ & Crew et al. (2005a)\\
896: GRB 050922C & CPL & $-0.83_{-0.23}^{+0.26}$ & - & $143 \pm 39$ & $1.37 \pm 0.06$ & Crew et al. (2005b)\\
897: GRB 051008 & CPL & $-0.975_{-0.078}^{+0.086}$ & - & $886 \pm 157$ & $1.13 \pm 0.05$ & Golenetskii et al. (2005e)\\
898: GRB 051109A & CPL & $-1.25_{-0.44}^{+0.59}$ & - & $224 \pm 141$ & $1.5 \pm 0.2$ & Golenetskii et al. (2005f)\\
899: GRB 060105 & CPL & $-0.83 \pm 0.03$ & - & $424_{-22}^{+25}$ & $1.07 \pm 0.04$ & Golenetskii et al. (2006a)\\
900: GRB 060117 & Band & $-1.52_{-0.07}^{+0.08}$ & $-2.9_{-0.5}^{+0.3}$ & $89 \pm 5$ & $1.93 \pm 0.03$ & Golenetskii et al. (2006b)\\
901: GRB 060313 & CPL & $-0.6 \pm 0.2$ & - & $922_{-177}^{+306}$ & $0.70 \pm 0.07$ & Golenetskii et al. (2006c)\\
902: GRB 060510 & CPL & $-1.66 \pm 0.07$ & - & $184_{-24}^{+36}$ & $1.57 \pm 0.07$ & Golenetskii et al. (2006d)\\
903: GRB 060813 & Band & $-0.53_{-0.14}^{+0.16}$ & $-2.6_{-0.5}^{+0.3}$ & $192_{-18}^{+20}$ & $1.36 \pm 0.04$ & Golenetskii et al. (2006e)\\
904: GRB 060814 & CPL & $-1.4 \pm 0.2$ & - & $257_{-58}^{+122}$ & $1.54 \pm 0.03$ & Golenetskii et al. (2006f)\\
905: GRB 060904A & Band & $-1.0 \pm 0.2$ & $-2.6_{-1.0}^{+0.4}$ & $163 \pm 31$ & $1.55 \pm 0.04$ & Golenetskii et al. (2006g)\\
906: GRB 061007 & Band & $-0.7 \pm 0.4$ & $-2.6_{-0.5}^{+0.3}$ & $399_{-18}^{+19}$ & $1.03 \pm 0.03$ & Golenetskii et al. (2006h)\\
907: GRB 061021 & CPL & $-1.2 \pm 0.1$ & - & $777_{-237}^{+549}$ & $1.30 \pm 0.06$ & Golenetskii et al. (2006i)\\
908: GRB 061121 & CPL & $-1.32 \pm 0.05$ & - & $606_{-72}^{+90}$ & $1.41 \pm 0.03$ & Golenetskii et al. (2006j)\\
909: GRB 061201 & CPL & $-0.36_{-0.65}^{+0.40}$ & - & $873_{-284}^{+458}$ & $0.8 \pm 0.1$ & Golenetskii et al. (2006k)\\
910: GRB 061222A & Band & $-0.94_{-0.13}^{+0.14}$ & $-2.4_{-1.2}^{+0.3}$ & $283_{-42}^{+59}$ & $1.35 \pm 0.04$ & Golenetskii et al. (2006l)\\
911: GRB 070220 & Band & $-1.2_{-0.2}^{+0.3}$ & $-2.0_{-0.4}^{+0.3}$ & $299_{-130}^{+204}$ & $1.40 \pm 0.04$ & Golenetskii et al. (2007a)\\
912: GRB 070328 & Band & $-1.0 \pm 0.1$ & $-2.0_{-0.4}^{+0.2}$ & $496_{-117}^{+172}$ & $1.24 \pm 0.04$ & Golenetskii et al. (2007b)\\
913: GRB 070420 & CPL & $-1.2 \pm 0.2$ & - & $147_{-19}^{+29}$ & $1.56 \pm 0.05$ & Golenetskii et al. (2007c)\\
914: GRB 070508 & CPL & $-0.81 \pm 0.07$ & - & $188 \pm 8$ & $1.36 \pm 0.03$ & Golenetskii et al. (2007d)\\
915: GRB 070521 & CPL & $-0.9 \pm 0.1$ & - & $222_{-21}^{+27}$ & $1.36 \pm 0.04$ & Golenetskii et al. (2007e)\\\hline
916: \end{tabular}}
917: }
918: \tablenotetext{a}{99\% upper limit}
919: \tablenotetext{b}{90\% upper limit}
920: \tablenotetext{c}{The confidence interval is not calculated because of $\chi^{2}_{\nu}$ $>$ 2.}
921: \end{table}
922:
923: %% figures
924: \newpage
925: \begin{figure}
926: \centerline{
927: \includegraphics[scale=0.6,angle=-90]{f1a.eps}}
928: \vspace{1cm}
929: \centerline{
930: \hspace{0.3cm}
931: \includegraphics[scale=0.6,angle=-90]{f1b.eps}}
932: \caption{Input parameters in the spectral simulations. The distribution of the low energy photon
933: index $\alpha$, the high energy photon index $\beta$ in the Band function, the low energy photon
934: index $\alpha$ in a CPL model from the BATSE GRB sample,
935: and the BAT T$_{100}$ duration from top to bottom, respectively. The dotted line represents the best fit in a
936: gaussian.}
937: \label{sim_input_paramters}
938: \end{figure}
939:
940: \newpage
941: % Ep-flux
942: \begin{figure}
943: \centerline{
944: \includegraphics[scale=0.5]{f2a.eps}
945: \hspace{-1.0cm}
946: \includegraphics[scale=0.5]{f2b.eps}}
947: \centerline{
948: \includegraphics[scale=0.5]{f2c.eps}
949: \hspace{-1.0cm}
950: \includegraphics[scale=0.5]{f2d.eps}}
951: \centerline{
952: \includegraphics[scale=0.5]{f2e.eps}
953: \hspace{-1.0cm}
954: \includegraphics[scale=0.5]{f2f.eps}}
955: \centerline{
956: \includegraphics[scale=0.5]{f2g.eps}
957: \hspace{-1.0cm}
958: \includegraphics[scale=0.5]{f2h.eps}}
959: \caption{The contour maps showing, as a function of the simulated $\ep$ and the
960: energy flux in the 15-150 keV band, the number of simulated spectra which have
961: $\Delta\chi^{2}$ $>$ 6 (left row for the Band function: $\Delta\chi^{2} \equiv \chi^{2}_{\rm PL} - \chi^{2}_{\rm Band}$;
962: right row for a CPL model: $\Delta\chi^{2} \equiv \chi^{2}_{\rm PL} - \chi^{2}_{\rm CPL}$). The incident angles of
963: the simulations are 0$^{\circ}$, 15$^{\circ}$, 30$^{\circ}$ and 50$^{\circ}$ from top to bottom.}
964: \label{fig:sim_ep_fluence}
965: \end{figure}
966:
967: \newpage
968: % explain difference in countour between Band and CPL
969: \begin{figure}
970: \centerline{
971: \includegraphics[scale=0.6,angle=-90]{f3.eps}}
972: \caption{The schematic drawing of the photon spectra of the Band
973: function (black) and a CPL model (red) with $\ep$ of 15 keV. The low
974: energy photon index is $-1$ for both models. The high energy photon
975: index of the Band function is $-2.3$. A PL model with a photon index
976: of $-2.3$ is also overlaid in the plot (blue). The vertical dotted lines
977: are the BAT observed energy band of 15-150 keV.}
978: \label{fig:exp_band_cpl_lowep}
979: \end{figure}
980:
981: \newpage
982: % histogram of PL vs. CPL with theta<25deg and theta>50deg
983: \begin{figure}
984: \centerline{
985: \includegraphics[scale=0.5,angle=-90]{f4a.eps}}
986: \vspace{0.5cm}
987: \centerline{
988: \includegraphics[scale=0.5,angle=-90]{f4b.eps}}
989: \caption{The number of GRBs acceptably fit by a PL model (top panel) and by a CPL
990: model (bottom panel) as a function of the fluence in the 15-150 keV band. The samples
991: of the incident angles of bursts less than 25 degrees (top) and larger than
992: 50 degrees (bottom).}
993: \label{fig:hist_pl_cpl}
994: \end{figure}
995:
996: \newpage
997: % Ep-Gamma relation
998: \begin{figure}
999: \centerline{
1000: \includegraphics[scale=0.5]{f5a.eps}
1001: \hspace{-1.0cm}
1002: \includegraphics[scale=0.5]{f5b.eps}}
1003: \centerline{
1004: \includegraphics[scale=0.5]{f5c.eps}
1005: \hspace{-1.0cm}
1006: \includegraphics[scale=0.5]{f5d.eps}}
1007: \centerline{
1008: \includegraphics[scale=0.5]{f5e.eps}
1009: \hspace{-1.0cm}
1010: \includegraphics[scale=0.5]{f5f.eps}}
1011: \centerline{
1012: \includegraphics[scale=0.5]{f5g.eps}
1013: \hspace{-1.0cm}
1014: \includegraphics[scale=0.5]{f5h.eps}}
1015: \caption{Contour maps showing the number of simulated spectra as a
1016: function of photon index and
1017: input $\ep$. The left and right rows are the Band function and
1018: a CPL model, respectively. The incident angles of the simulations
1019: are 0$^{\circ}$, 15$^{\circ}$, 30$^{\circ}$, and 50$^{\circ}$
1020: from top to bottom.}
1021: \label{fig:sim_phindex}
1022: \end{figure}
1023:
1024:
1025:
1026: \newpage
1027: % Ep-Gamma relation fitting
1028: \begin{figure}
1029: \centerline{
1030: \includegraphics[scale=0.5,angle=-90]{f6a.eps}}
1031: \vspace{1cm}
1032: \centerline{
1033: \includegraphics[scale=0.5,angle=-90]{f6b.eps}}
1034: \caption{The best fit $\ep$ - $\Gamma$ relations (solid line) and the
1035: lower and higher 1-$\sigma$ confidence level of the relations (dashed
1036: lines) for the Band function (top) and a CPL model (bottom) with the
1037: data points (circles: $\ep$ at the peak of the histogram of
1038: $\Gamma$, squares: $\ep$ value of 16\% crossing point of the histogram
1039: of $\Gamma$, and triangles: $\ep$ value of 84\% crossing point of the histogram
1040: of $\Gamma$). The black, red, green and blue show the cases of
1041: incident angles 0$^{\circ}$, 15$^{\circ}$, 30$^{\circ}$ and
1042: 50$^{\circ}$, respectively.}
1043: \label{fig:ep_gamma_fit}
1044: \end{figure}
1045:
1046: \newpage
1047: % Theta distribution
1048: \begin{figure}
1049: \centerline{
1050: \includegraphics[scale=0.6,angle=-90]{f7.eps}}
1051: \caption{The incident angle ($\theta$) distribution of the BAT GRBs.
1052: The dotted line is the best fit gaussian model.}
1053: \label{fig:theta}
1054: \end{figure}
1055:
1056: \newpage
1057: % Theta weighted Ep-Gamma
1058: \begin{figure}
1059: \centerline{
1060: \includegraphics[scale=0.9]{f8a.eps}}
1061: \vspace{1cm}
1062: \centerline{
1063: \includegraphics[scale=0.9]{f8b.eps}}
1064: \caption{Contour maps showing the number of simulated spectra as a
1065: function of photon index and input $\ep$ after weighting the
1066: simulation results of 0$^{\circ}$, 15$^{\circ}$, 30$^{\circ}$,
1067: and 50$^{\circ}$ incident angles by the incident angle distribution
1068: of the BAT GRBs shown in Figure \ref{fig:theta} (top: the Band function and bottom:
1069: a CPL model).}
1070: \label{fig:weighted_ep_gamma}
1071: \end{figure}
1072:
1073: \newpage
1074: % Theta weighted Ep-Gamma fitting
1075: \begin{figure}
1076: \centerline{
1077: \includegraphics[scale=0.5,angle=-90]{f9a.eps}}
1078: \vspace{1cm}
1079: \centerline{
1080: \includegraphics[scale=0.5,angle=-90]{f9b.eps}}
1081: \caption{The best fit weighted $\ep$ - $\Gamma$ relations by the
1082: incident angles (solid line) and the
1083: lower and higher 1-$\sigma$ confidence level of the relations (dashed
1084: lines) for the Band function (top) and a CPL model (bottom) with the
1085: data points (circles: $\ep$ at the peak of the histogram of
1086: $\Gamma$, squares: $\ep$ value of 16\% crossing point of the histogram
1087: of $\Gamma$, and triangles: $\ep$ value of 84\% crossing point of the histogram
1088: of $\Gamma$).}
1089: \label{fig:weighted_ep_gamma_fit}
1090: \end{figure}
1091:
1092: \newpage
1093: % Compare with observations
1094: \begin{figure}
1095: \centerline{
1096: \includegraphics[scale=0.5,angle=0]{f10a.eps}}
1097: \vspace{1cm}
1098: \centerline{
1099: \includegraphics[scale=0.5,angle=0]{f10b.eps}}
1100: \caption{The distribution of $\ep$ and photon index, $\Gamma$, in
1101: a PL fit in the BAT GRB sample (black circles). The GRBs which were
1102: simultaneously observed by {\it HETE-2} and {\it Konus-Wind} are
1103: overlaid (blue triangles). The weighted $\ep$ - $\Gamma$ relation
1104: for the Band function (top) and a CPL model (bottom) with 1-$\sigma$
1105: confidence level is overlaid on the data. {\it Inset}: The extended
1106: figures of $\Gamma$ from 1.3 to 2.3 where the $\ep$ - $\Gamma$ relation
1107: is valid.}
1108: \label{fig:comp_epeak_gamma_obs}
1109: \end{figure}
1110:
1111: \newpage
1112: % Compare Ep between ep-gamma and Nat
1113: \begin{figure}
1114: \centerline{
1115: \includegraphics[scale=1.0,angle=0]{f11.eps}}
1116: \caption{The relationship between $\ep$ reported by \citet{butler2007}
1117: and $\ep$ derived from the weighted $\ep$ - $\Gamma$ relation for the
1118: Band function. The sample only contains long bursts which have a PL
1119: photon index $\Gamma$ from 1.3 to 2.3. Both $\ep$ distributions of
1120: \citet{butler2007} and the $\ep$ - $\Gamma$ relation for the same sample
1121: are shown in the histograms.}
1122: \label{fig:comp_taka_nat_ep}
1123: \end{figure}
1124:
1125: %% User Guide for details.
1126: %%
1127: %% No more than seven \figcaption commands are allowed per page,
1128: %% so if you have more than seven captions, insert a \clearpage
1129: %% after every seventh one.
1130:
1131: %% Tables should be submitted one per page, so put a \clearpage before
1132: %% each one.
1133:
1134: %% Two options are available to the author for producing tables: the
1135: %% deluxetable environment provided by the AASTeX package or the LaTeX
1136: %% table environment. Use of deluxetable is preferred.
1137: %%
1138:
1139: %% Three table samples follow, two marked up in the deluxetable environment,
1140: %% one marked up as a LaTeX table.
1141:
1142: %% In this first example, note that the \tabletypesize{}
1143: %% command has been used to reduce the font size of the table.
1144: %% We also use the \rotate command to rotate the table to
1145: %% landscape orientation since it is very wide even at the
1146: %% reduced font size.
1147: %%
1148: %% Note also that the \label command needs to be placed
1149: %% inside the \tablecaption.
1150:
1151: %% This table also includes a table comment indicating that the full
1152: %% version will be available in machine-readable format in the electronic
1153: %% edition.
1154: %%
1155: \clearpage
1156:
1157: \end{document}
1158: