1:
2: \documentclass[12pt,preprint]{aastex}
3:
4: % for a referee version
5:
6: % \documentclass[manuscript]{aastex}
7:
8: %
9: %documentclass[article]{aa}
10: \usepackage{epsfig}
11: %
12:
13: \begin{document}
14:
15: \newcommand{\gsim}{\hbox{\rlap{$^>$}$_\sim$}}
16:
17: \title{On the X-ray lines in the afterglows of GRBs}
18:
19: \author{Shlomo Dado\altaffilmark{1}, Arnon Dar\altaffilmark{1} and
20: A. De R\'ujula\altaffilmark{2}}
21:
22: \altaffiltext{1}{Physics Department and Space Research Institute,
23: Technion, Haifa 32000, Israel}
24: \altaffiltext{2}{Theory Division, CERN,CH-1211 Geneva 23, Switzerland}
25:
26: % \maketitle
27:
28: \begin{abstract}
29: The observation of X-ray lines in the afterglow of GRB 011211 has been
30: reported, and challenged. The lines were interpreted as blue-shifted X-rays
31: characteristic of a set of photoionized ``metals'', located
32: in a section of a supernova shell illuminated by a GRB emitted a couple
33: of days after the supernova explosion. We show that the most prominent
34: reported lines coincide with the ones predicted in the ``cannonball''
35: model of GRBs. In this model, the putative signatures are Hydrogen lines,
36: boosted by the (highly relativistic) motion of the cannonballs (CBs). The
37: corresponding Doppler boost can be extracted from the fit to the observed
38: I-, R- and V-band light-curves of the optical afterglow of GRB 011211, so
39: that, since the redshift is also known, the line energies are --in the CB
40: model-- predicted. We also discuss other GRBs of known redshift which show
41: spectral features generally interpreted as Fe lines, or Fe recombination
42: edges. The ensemble of results is very encouraging from the CB-model's
43: point of view, but the data on each individual GRB are not good enough to
44: draw (any) objectively decisive conclusions. We outline a strategy for
45: X-ray observers to search for lines which, in the CB model, move
46: predictably from higher to lower energies.
47:
48: \end{abstract}
49:
50: \keywords{gamma rays: bursts---X rays: Lines}
51:
52:
53:
54: \section*{Introduction}
55:
56: There is mounting evidence from late-time observations of the optical
57: afterglows (AGs) of relatively nearby (redshift $\rm z<1$) gamma ray bursts
58: (GRBs) that long duration GRBs are produced in the explosions of
59: supernovae akin to
60: SN1998bw (Galama et al. 1998), by the ejection of ordinary baryonic
61: matter --essentially ionized Hydrogen-- in the form of plasmoids or
62: ``cannonballs'' (CBs), with very highly
63: relativistic Lorentz factors ($\gamma\sim 10^3$)
64: (Dar and De R\'ujula 2000a,b, Dado et al. 2002a,b,c),
65: but otherwise
66: similar to the ones observed in quasars (Marscher et al. 2002) and
67: microquasars (e.g., Mirabel and Rodriguez 1994; Belloni et al. 1997;
68: Mirabel and Rodriguez 1999; Rodriguez and Mirabel 1999 and references
69: therein). The ejection of these cannonballs (CBs) close to the
70: line of sight makes their sky-projected motion appear extremely
71: superluminal\footnote{With the exception (Dar and De R\'ujula 2000a)
72: of the very close-by
73: GRB 980425, associated to SN1998bw, this ``hyperluminal'' motion is
74: not directly observable, but it was suggested (Dado et al. 2002d) that it
75: gives
76: rise to the observed scintillations in the GRB radio afterglows (Taylor et al.
77: 1997) and may thereby be measurable.}.
78:
79: On Dec. 11, 19:09:21 UT 2001 the long duration ($\sim 270$ s)
80: GRB 011211 was detected in the constellation Crater by BeppoSAX (Gandolfi
81: et al. 2001). Approximately ten hours after the GRB, its optical afterglow
82: was detected by Grav et al. (2001) and was followed by measurements of its
83: declining light-curve by Bloom et al. (2001), Jensen et al. (2001),
84: Holland et al. (2002), Bhargavi et al. 2001, Fiore et al. (2001), Burud et al. (2001)
85: Covino et al. (2001); and Fruchter et al. (2001), who
86: also measured its redshift: $\rm z=2.141$, confirmed in turn by Gladders et al.
87: (2001). The GRB's host galaxy was detected, with a red magnitude
88: $ \rm R_{host}=25.0\pm 0.3$, by Burud et al. (2001).
89:
90: Observations with XMM-Newton of the X-ray afterglow of GRB 011211 started
91: at 06:16:56 UT on December 12, 2001, 11 h (39.6 ks) after burst, and lasted 27 ks
92: (Reeves et al. 2002a). The analysis of the X-ray spectrum revealed significant
93: evidence for emission lines only in the first 10 ks of observations. The
94: emission lines that were fitted to the first 5 ks data had energies of
95: $\rm 1.40 \pm 0.05\,keV$, $\rm 2.19 \pm 0.04\,keV$, $\rm 2.81 \pm
96: 0.04\,keV$, $\rm 3.79 \pm 0.07\,keV$, and $\rm 4.51 \pm 0.12\,keV$,
97: in the GRB rest frame. They were interpreted by the observers as $\rm
98: K\alpha$ lines from MgXI, SiXIV, SXVI, ArXVIII and CaXX,
99: blueshifted by the motion at $\rm \beta= v/c= 0.086\pm 0.04$
100: of a shell ejected by a massive GRB progenitor in a supernova (SN) explosion
101: having occurred a couple of days prior to the GRB.
102: In this interpretation, a section of the SN shell near the line of
103: sight was illuminated and reheated to a temperature of $\rm T \sim 4.5\pm
104: 0.5 $ keV by the beamed GRB, and it emitted the blueshifted X-ray lines.
105:
106: Borodzin and Trudolyubov (2002) have criticized the above interpretation
107: by noticing that the data showing the lines were accumulated during
108: the first 5 ks of observations, while the source was located near the
109: edge of a CCD chip, and that the lines disappeared as the satellite was
110: subsequently
111: repositioned. Moreover, the background data collected over the
112: edge of the CCD show a very significant peak at the position of
113: the most prominent alleged line. Rutledge and Sako (2002) have also
114: criticized the significance of these data on statistical grounds.
115: Reeves et al. (2002b) have responded to these critiques and
116: insisted on the significance of their results, though they find
117: that a fit without the two lines of minimum and maximum energy
118: (attributed to MgXI and CaXX) is as good, or even a little better,
119: than the fit with all five lines.
120:
121: We cannot enter into the above controversy. In what follows,
122: we discuss the data in Reeves et al. (2002a) at face value for, even if
123: their significance is weakened, they constitute a good stage
124: within which to discuss the predictions of the CB model
125: (Dar and De R\'ujula 2000a) concerning
126: X-ray lines in GRB afterglows. We concentrate on GRB 011211 because
127: it has, so far, the best measured X-ray spectrum, but we also discuss
128: other GRBs of known redshift in which $\rm Fe$ lines and/or a
129: recombination edge have been claimed to be observed (GRB 970508: Piro et
130: al. 1998; GRB 970828: Yoshida et al. 1999; 2001 and GRB 991216: Piro et
131: al. 2000).
132: In all cases these ``lines'' or ``recombination edges'' are not truly
133: observationally established; we shall often refer to them as
134: spectral {\it features,} for the sake of precision.
135:
136: The subject of X-ray lines in GRB
137: afterglows has attracted considerable theoretical attention:
138: Bottcher (2000),
139: Bottcher and Fryer (2001),
140: Ghisellini et al. (2002),
141: Kumar and Narajan (2002),
142: Lazzati et al. (1999, 2002),
143: Meszaros and Rees (2001),
144: Rees and Meszaros (2000),
145: Vietri et al. (2001),
146: Wang et al. (2002),
147: Weth et al. (2000).
148:
149: The interpretation of the X-ray features in Reeves et al. (2002a)
150: as metal lines is not without
151: problems. First, the non-detection of the Fe $\rm K\alpha$ line was
152: argued to be due to the relatively long time it takes the $\beta$ decay
153: chain $\rm Ni^{56}\rightarrow Co^{56}\rightarrow Fe^{56}$ to produce $\rm
154: Fe^{56}$ in supernova explosions. This appears to be inconsistent with
155: the fact that the only X-ray lines with large flux and equivalent width
156: previously claimed to be detected in GRB afterglows were attributed to Fe lines:
157: the BeppoSAX results for GRB 970508 (Piro et al. 1998) and GRB 000214
158: (Antonelli et al. 2000), the ASCA observations for GRB 970828 (Yoshida et al.
159: 1999; 2001) and the Chandra data on GRB 991216 (Piro et al. 2000).
160: Second, the fitted blueshift of the X-ray lines is supposedly due to the
161: beaming of the GRB radiation that illuminates only a small section of the
162: expanding SN shell, near the line of sight. The energy deposition time in
163: such a segment is very short {\it in the
164: observer frame}: the GRB ejecta is moving initially with a large Lorentz
165: factor $\gamma$, and it overtakes a SN shell with an
166: estimated radius $\rm R_{SNS}\sim 10^{15}$ cm in $\rm t\sim (1+z)\,
167: R_{SNS}/c\, \gamma^2\leq 10$ sec of observer's time, for $\gamma\! >\! 100$.
168: The radiative cooling time of an
169: optically thin SN shell with an electron density $\rm n_e \sim 10^{15}\,
170: cm^{-3}$ and a temperature of 4.5 keV
171: is also extremely short: $\tau\ll 1$ s. The arrival times
172: (in the observer frame) of recombination photons from the SN shell sector
173: illuminated by a GRB jet of opening angle $\theta=20^o$ are spread over
174: $\rm t=R_{SNS}\, (1+z)\, (1-cos\theta)/c=1.75$ h after the GRB, while the
175: putative XMM lines were observed 11 h after burst. These, and other puzzling
176: geometrical and physical details of the model, leave ample room for other
177: interpretations of the observations, should they be real.
178:
179: An alternative interpretation (Dar and De R\'ujula
180: 2001a) is that the spectral features observed in the
181: X-ray afterglows of GRBs are optical Hydrogen-recombination lines from
182: the CBs that produce GRBs, Doppler shifted to the X-ray
183: band by the CBs' highly relativistic motion ($\gamma\sim 10^3$) and
184: observed\footnote{Doppler-shifted Lyman, Balmer and
185: HeI lines have been detected from the mildly relativistic jets of
186: SS433 (e.g. Margon 1984; Kotani et al. 1996; Eikenberry et al. 2001).}
187: at very small angles ($\theta\sim 1/\gamma$).
188:
189: In this paper we show that the energies of the X-ray emission lines perhaps
190: detected in the afterglow of GRB 011211 happen to coincide with the
191: energies of Hydrogen's Balmer and Lyman lines,
192: redshifted by $\rm 1+z=3.141$ due to the cosmic expansion, and
193: blueshifted by a Doppler factor $\delta\sim 835$. This value of $\delta$
194: is, as we shall see, that of the CBs at the time of the observation of
195: the putative X-ray lines (some 11h after burst)
196: obtained from a Cannonball-Model fit to the light-curve of the optical
197: afterglow of GRB 011211, prior to the reported X-ray observations. What
198: this means is that the positions of the X-ray lines ---{\it predicted} in the
199: CB model--- coincide with the observed positions of the most
200: prominent spectral features in the data.
201:
202: In discussing the previous indications for ``Fe'' lines in the AGs of
203: GRBs 970508, 970828 and 991216
204: (Dar and De R\'ujula 2001a) we argued that they were
205: compatible with $\rm Ly\alpha$ emission, though the data were
206: insufficient to make a decisive distinction between a highly boosted
207: hydrogen line and a merely redshifted Fe line. Here, we
208: rediscuss this issue in more detail, now that we also
209: have independent a-priori determinations of the values of
210: $\delta$ in each individual GRB at the time of the corresponding X-ray
211: observations. As for the case
212: of GRB 011211, all prominent features in the spectra coincide in
213: energy with lines that can be expected in the CB model.
214:
215: The CB model predicts that the X-ray lines should be relatively
216: narrow and move in
217: time from higher to lower frequencies, as the CBs decelerate
218: while ploughing though the interstellar medium (ISM). The blending of
219: the emissions from unresolved CBs with somewhat different Doppler factors,
220: and/or a poor energy resolution, may broaden the lines considerably and
221: conceal the time-dependence of their energy. In the current data, the
222: limited energy resolution and the required integration
223: over relatively long time-intervals would certainly have precluded
224: the observation of the predicted line motion.
225:
226: \section*{CB model fit to the AG of GRB 011211}
227: \noindent
228: We do not give here a detailed description of the CB model, which we have
229: discussed {\it at nauseam} elsewhere (Dar and De R\'ujula 2000a,b, Dado et al.
230: 2002a,d). We simply reproduce the formulae required for the
231: analysis at hand.
232:
233: In the CB model the afterglow has three origins: the ejected CBs, the
234: concomitant SN explosion, and the host galaxy (HG). These components are
235: usually unresolved in the measured GRB afterglow, so that the
236: corresponding light curves and spectra are measures of the cumulative energy
237: flux density:
238: \begin{equation}
239: \rm F_{AG}=F_{CBs}+F_{SN}+F_{HG}\, .
240: \label{sum}
241: \end{equation}
242: The contribution from the host galaxy dominates
243: the light-curve at late times and was fitted (in each band) to the late
244: afterglow. The contribution of the supernova was modelled by assuming an
245: SN1998bw-like contribution placed at the GRB redshift, $\rm z=2.141$.
246: In this particular GRB, as in all the ones with redshift $\rm z>1.2$, the
247: SN contribution is too dim to be observable (Dado et al. 2002a).
248:
249: In the CB model the jetted cannonballs are made of ordinary matter, mainly
250: hydrogen. For the first $\sim 10^3$ seconds of observer's time, a CB is still
251: cooling fast and emitting via thermal bremsstrahlung (Dado et al. 2002a),
252: but after
253: that its emissivity is dominated by synchrotron emission from ISM
254: electrons that penetrate in it. Integrated over frequency, this synchrotron
255: emissivity is approximately equal to the energy deposition rate of the
256: ISM electrons in
257: the CB\footnote{The kinetic energy of a CB is mainly lost to the ISM
258: protons it scatters; only a fraction $\rm\leq m_e/m_p$ is re-emitted by
259: electrons, as the AG.}. The electrons from the ISM that enter the CBs are Fermi
260: accelerated there to a broken power-law energy distribution with a ``break''
261: energy (or more appropriately a ``bend'' energy) equal to their incident energy
262: in the CBs' rest frame. In that frame, the electrons' synchrotron emission
263: (prior to attenuation corrections)
264: has an approximate spectral energy density (Dado et al. 2002d):
265: \begin{equation}
266: \rm F_{_{CB}}[\nu,t] = E_\gamma\,{dn_\gamma\over d\,E_\gamma}\sim f_0\;
267: { (p-2)\, \gamma^2\, \over (p-1)\,\nu_b}{ [\nu /\nu_b]^{-1/2}
268: \over \sqrt{1+[\nu/ \nu_b]^{(p-1)}}}
269: \label{sync}
270: \end{equation}
271: where $\rm p\approx 2.2$ is the spectral index of the Fermi accelerated
272: electrons prior to the inclusion of radiation losses, $\rm f_0$ is an
273: explicit normalization constant proportional to the ISM baryon density
274: $\rm n_p$,
275: $\rm \gamma(t)=1/\sqrt{1-\beta^2}$ (with $\rm \beta=v/c$) is
276: the Lorentz factor of the CBs, and
277: $\rm \nu_b \simeq 1.87\times 10^{3}\,[\gamma(t)]^3\,
278: [n_p/10^{-3}cm^{-3}]^{1/2}$ Hz
279: is the ``injection bend'' frequency in the CB rest frame\footnote{This bend
280: frequency does not correspond to the conventional
281: synchrotron ``cooling break''. It is produced by an {\it injection bend}
282: in the high energy electron spectrum in the CB at the energy $\rm E_b =
283: \rm\gamma(t)\ m_e\, c^2$ with which the ISM electrons enter the CB
284: at a particular time in its decelerated motion (Dado et al. 2002d).}.
285: The X-ray frequency domain in Eq.~(\ref{sync}) is always at
286: $\nu\gg\nu_b$, so that the expected spectrum is
287: $\rm d\,n_\gamma/d\,E_\gamma
288: \approx E^{-\alpha}$, with a slope $\rm\alpha=(p+2)/2\simeq 2.1$.
289: The radiation emitted by a CB is Doppler-shifted
290: and forward-collimated by its highly relativistic
291: motion, and redshifted by the cosmological expansion.
292: A distant observer sees a spectral energy flux:
293: \begin{equation}
294: \rm F_{obs}[\nu,t]\simeq
295: \rm { (1+z)\,\delta(t)^3\, R^2\, A(\nu,z)\,\over D_L^2}\,
296: F_{_{CB}}\left[{(1+z)\,\nu\over\delta(t)},{\delta(t)\,t\over 1+z}
297: \right]\, ,
298: \label{Fnuobser}
299: \end{equation}
300: where $\rm R$ is the radius of the CB (which
301: in the CB model tends to a calculable constant value
302: ${\rm R_{max}}={\cal{O}}(10^{14})$ cm, in minutes of observer's time),
303: $\rm A(\nu,z)$ is the total extinction along the line of sight to the GRB,
304: $\rm D_L(z)$ is the luminosity
305: distance\footnote{The cosmological parameters we use are: $\rm H_0=65$
306: km/(s Mpc), ${\rm \Omega_M}=0.3$ and ${\rm \Omega_\Lambda}=0.7$.} and
307: $\rm\delta(t)$ is the Doppler factor of the light emitted by the CB:
308: \begin{equation}
309: \rm
310: \delta(t)=
311: {1\over\gamma(t)\,(1-\beta(t)\cos\theta)}
312: \simeq{2\, \gamma(t)\over 1+\theta^2\gamma(t)^2}\; ,
313: \label{doppler}
314: \end{equation}
315: where $\theta$ is the angle between the CB's direction of motion
316: and the line of sight to the observer.
317: The last approximation is valid in the domain
318: of interest for GRBs: $\rm \gamma^2\gg 1$ and $\theta^2\ll 1$.
319: The total AG is the sum over CBs (or large individual GRB pulses)
320: of the flux of Eq.~(\ref{Fnuobser}).
321:
322: For an interstellar medium of constant baryon density $\rm n_p$, the
323: deceleration of the CBs results in a Lorentz factor, $\rm\gamma(t)$, that is
324: given by (Dado et al. 2002a):
325: \begin{eqnarray}
326: \rm \gamma&=&\rm\gamma(\gamma_0,\theta,x_\infty;t)
327: =\rm {1\over B} \,\left[\theta^2+C\,\theta^4+{1\over C}\right]\; ,\nonumber\\
328: \rm C&\equiv&\rm
329: \left[{2\over B^2+2\,\theta^6+B\,\sqrt{B^2+4\,\theta^6}}\right]^{1/3}\; ,
330: \nonumber\\
331: \rm B&\equiv&\rm
332: {1\over \gamma_0^3}+{3\,\theta^2\over\gamma_0}+
333: {6\,c\, t\over (1+z)\, x_\infty}\; ,
334: \label{cubic}
335: \end{eqnarray}
336: where $\gamma_0=\gamma(0)$, and $\rm x_\infty=N_{_{CB}}/(\pi\, R_{max}^2\, n_p)$
337: characterizes the CB's slow-down in terms of
338: $\rm N_{_{CB}}$, its baryon number and $\rm R_{max}$,
339: its radius (it takes a distance $\rm x_\infty/\gamma_0$, typically of
340: ${\cal{O}}(1)$ kp, for
341: the CB to slow down to half its original Lorentz factor).
342:
343: In Fig.~(\ref{ag011211}), we show that the
344: optical afterglow of GRB 011211 in the IRV bands is very well fitted in
345: the cannonball model of GRBs, by use of Eqs.~(\ref{sync}) to (\ref{cubic}).
346: Besides the overall normalization, the fit involves three parameters:
347: $\rm \theta=1.159\pm 0.005$ mrad; $\gamma_0=824\pm 2$, and
348: $\rm x_\infty=0.271\pm 0.004$ Mpc.
349: We have fixed the spectral index
350: in Eq.~(\ref{sync}) to
351: $\rm p=2.2$, since that value is compatible with the one fit,
352: in the same manner, to {\it all} of the optical, X-ray and radio AG light-curves
353: of GRBs of known redshift (Dado et al. 2002a,d).
354: There is a reason why in this fit we have not used the X-ray light curve
355: as part of the input. The contribution of the lines is a significant fraction
356: of the X-ray count rate, and they fade away rapidly in the observational
357: period between 0.46 and 0.77 days after burst. The lines are, in the CB model, a
358: contribution that adds to the synchrotron-radiation described in
359: Eqs.~(\ref{sync}) to (\ref{cubic}). The X-ray light curve, thus, should
360: decline faster that its synchrotron component. The decline observed
361: in the measured interval is $\sim 50$\%, while the prediction is 32\% for
362: the synchrotron component, as determined from the fit to the optical
363: light curves.
364:
365:
366: It is consuetudinary in the GRB afterglow literature to quote values
367: of $\chi^2$ for the fits, and errors for the parameters. Yet, the
368: models are not fundamental theories, but rough approximations
369: of no doubt hideously complicated phenomena. Moreover, the dominant
370: contribution to the $\chi^2$ values very often originates in the spread
371: of almost simultaneous measurements; Fig.~(\ref{ag011211}) shows
372: this to be the case for GRB 011211, whose $\chi^2$/d.o.f. in our
373: CB-model fit is 1.4 for 26 data points\footnote{We have dealt with
374: the problem of partial data incompatibility with the method recommended
375: by the Particle Data Group (2000), tantamount in the cases at hand to
376: equating the difference between the extreme central values to a
377: formal $\sim 2\,\sigma$ spread.}. In this fit, we approximate
378: the ISM-density and the CB radius by constants. The tiny nominal
379: statistical errors of the fit, given in the previous paragraph, do
380: not include the systematic effects of deviations from these
381: approximations, nor do they reflect the systematic errors in the
382: data, involving the use of different detectors, assumptions about
383: absorption, etc. Thus, even though our results here and elsewhere
384: are very good, it would be misleading to emphasize the ${\cal{O}}$
385: (1\%) high
386: ``precision'' of our predicted line energies, since there is simply
387: no way to know what the ``true'' errors in the parameters are.
388: For other GRBs, we will not report the nominal parameter errors,
389: which are also minute.
390:
391: \section*{The X-ray line data of GRB 011211}
392: \noindent
393: The X-ray spectrum of this GRB has been well measured, in comparison
394: with previous cases. This is shown in Fig~(\ref{211spectrum}), which
395: we have borrowed from the data analysis by Borodzin and Trudolyubov
396: (2002). These authors find that the spectrum is compatible with
397: a power-law of slope $\alpha=2.14\pm 0.03$, modified only by absorption in the
398: Galaxy. A slope $\sim 2.1$ is expected in the CB model and
399: ---as extracted from the time-dependence of the optical and/or X-ray
400: AGs--- it is compatible with the observations for all GRBs of known
401: redshift (Dado et al. 2002a,b,c). In Fig~(\ref{211spectrum})
402: we also show how very compatible with the data at hand the expected
403: $\alpha \approx 2.1$ actually is.
404:
405: The observation and the properties of the lines reported in Reeves
406: et al.~(2002a) are model dependent; they are in particular very
407: sensitive to the continuum underlying the peaks in the data. This
408: can be seen in Fig.~(\ref{211lines}a), where we have redrawn the
409: data in Fig.~2 of Reeves et al. (2002a) without a model curve to
410: guide the eye (this figure reports data for the 5 ks interval in
411: which the lines were seen). Quite clearly, one can draw a smooth
412: continuum on this figure, above which the alleged lines would lose
413: much of their significance. To draw such a continuum with as little
414: prejudice as possible we have first made a smooth fit to Fig.~1 of
415: Reeves at al.~(2002a), which displays the data on the complete 27
416: ks of observational time, where there are no significant line
417: features. We have then redrawn
418: this continuum\footnote{The zig-zag feature around 0.5 keV, also
419: present in Fig. 1 of Reeves et al., (2002a) is presumably the effect
420: of the oxygen absorption edge in our Galaxy, somewhat smoothed by
421: resolution.} on top of the 5 ks data, with a normalization meant
422: to underemphasize the possible non-smooth deviations; the result
423: is shown in Fig.~(\ref{211lines}b).
424: This procedure may not be a sophisticated data analysis,
425: but it is a sure way to account for detector-response and other
426: systematic effects, such as the energy-dependence of the detector's
427: effective area (all we are assuming is that these effects are the
428: same for the 5 ks and the 27 ks data sets).
429: Clearly, with the continuum ``background'' of Fig.~(\ref{211lines}b),
430: the two alleged lines at $\sim1.21$ and $\sim1.44$
431: keV are not significant (these are the putative ArXVIII and CaXX
432: lines). Also, the other three smaller-energy lines, particularly
433: the lowest-energy one,
434: are not very prominent\footnote{Recall that Reeves et al. (2002b)
435: also find that the inclusion of the lowest and highest energy lines
436: does not improve their fits.}. The vertical lines in
437: Fig.~(\ref{211lines}b) are CB model expectations, which we proceed
438: to discuss.
439:
440:
441:
442:
443: \section*{Line emission in the CB model}
444: \noindent
445:
446: As a CB ---in a time of ${\cal{O}}(1)$ s after it exits the transparent
447: outskirts of the shell of the SN associated with it--- becomes transparent
448: to the bulk of its enclosed radiation, its internal
449: radiation pressure drops abruptly and its transverse expansion rate
450: is quenched by collisionless, magnetic-field-mediated
451: interactions with the ISM (Dado et al. 2002a).
452: During this phase, the ISM electrons that enter the CB cool mainly by
453: synchrotron emission. The synchrotron emission is
454: partially reabsorbed by
455: the partially ionized CB through free-free transitions
456: at low radio frequencies and by bound--free and
457: bound--bound transitions at optical frequencies (in the CB rest frame).
458: The CB plasma cools mainly by line emission from
459: electron--proton recombinations.
460:
461: \subsection*{Line energies: the case of GRB 011211}
462: \noindent
463: At a given time $\rm t$, the CBs are viewed with a blue-shifting
464: Doppler factor $\rm \delta(t)$, so that a line of laboratory wavelength
465: $\rm \lambda_i$ would be observed at a redshift $\rm z$ to have an energy
466: uplifted by a ``boost'' factor $\rm B(t)$:
467: \begin{eqnarray}
468: \rm E_i(t)&=&\rm B(t) \, E_i^{lab}=B(t)\;{h\,c\over \lambda_i}\nonumber\\
469: \rm B(t)&\equiv&\rm{\delta(t)\over 1+z}\; ,
470: \label{Lineenergy}
471: \end{eqnarray}
472: with $\rm\delta(t)$ as in Eq.~(\ref{doppler}).
473: The parameters that we fit to the optical AG of GRB 011211, substituted
474: in Eqs.~(\ref{doppler},\ref{cubic}), result in $\rm \gamma(t)=668$
475: at $\rm t\sim 11$ h, the
476: time when the lines were seen. For a redshift $\rm z=2.141$, Eq.~(\ref{doppler})
477: implies $\rm \delta(t)\simeq 835$, so that lines at rest in the CBs would
478: be uplifted in energy by a factor $\rm B(t)\simeq 266$ at the time
479: of the observations.
480:
481: In Fig.~(\ref{linemigration}) we show the predicted evolution of the
482: energy $\rm E_i(t)\propto\delta(t)$ of the $\rm H$ lines as a function of
483: time, for the case of GRB 011211. The X-ray observations of Reeves et al.
484: (2002a) lasted too little for the line motion to have an observable effect,
485: given their limited statistics and energy resolution. This is also the
486: case for all the other GRBs to be discussed below.
487:
488: In normal dense astrophysical plasmas, e.g., plasma clouds in the
489: broad-line region of quasars (see, e.g., Laor et al. 1997
490: and references therein), the prominent Hydrogen lines are:
491: $\rm H\alpha [\lambda 6563]$,
492: $\rm H\beta [\lambda 4861]$, the higher energy Balmer lines accumulating
493: at $\rm H\infty [\lambda 3647]$,
494: and the $\rm Ly\alpha[\lambda 1215.7]$ line.
495: The first three of these lines are, as one can see in Fig.~(\ref{211lines}b),
496: at the positions where there are, perhaps, indications in the data of an excess
497: over a smooth continuum: the predicted energies are 0.50, 0.68 and 0.91 keV,
498: while the fit of Reeves et al. (2002a) results in $0.45\pm 0.05$, $0.70\pm 0.02$
499: and $0.89\pm 0.01$ keV, respectively.
500: The $\rm Ly\alpha$ line ought to be uplifted in energy
501: to $\sim 2.74$ keV, above the range shown in Fig.~(\ref{211lines}). Interestingly,
502: there is a feature in the 27 ks data (Fig. 1 of Reeves et al. 2002a) which,
503: although it is also not very significant, sits at that very
504: point\footnote{We do not have access to
505: that figure in an e-friendly format, and it
506: is too complicated to reproduce by hand, or by scanning.}.
507: All these features have widths comparable with the experimental
508: resolution of somewhat less than 100 eV.
509:
510: We have argued in Dar and De R\'ujula (2000, 2001a,b) and Dar et al.
511: (2000) that the ordinary matter
512: constituting cannonballs ought to be shattered by their violent
513: collision with the SN shell, and exit it in the form of unbound baryons
514: and electrons, so that the expected X-ray lines would be merely
515: hydrogenic. But it is quite possible that the collisions be somewhat
516: ``cushioned'' (Hubbard and Ferry, in preparation, Hubbard 2002) such
517: as to leave some
518: nuclei unscathed, as the CBs gather SN shell material in their passage:
519: in their collisions with shell nuclei, the baryons or nuclei of the CBs
520: lose a considerable fraction of their initial Lorentz factor
521: (Dar and De R\'ujula 2001b), but they
522: may do it in many soft collisions, as opposed to a few hard ones.
523: It that case, one may expect to see also $\rm He$- or even
524: ``metal'' lines, uplifted in energy by as much as the $\rm H$ lines are.
525: The predicted position of the $\rm He\alpha[\lambda 5875]$
526: of $\rm HeI$
527: is also shown in Fig.~(\ref{211lines}b).
528: One quasi-degenerate
529: example of lines, prominent in the broad line region of quasars, is the pair
530: $\rm MgII[\lambda 2796.3;\, \lambda 2803.5]$, which in the case
531: at hand should appear at 1.19 keV, see Fig.~(\ref{211lines}b).
532: This is where Reeves et al. (2002a)
533: claim to see Ar XVIII, at $1.21\pm 0.02$ keV.
534: But the possible choices (other than for $\rm H$ and
535: perhaps $\rm He$ and $\rm Ni$ lines) are far too vast to
536: draw definite conclusions from these very scant data.
537: This is even more so in the ``standard''
538: interpretation, in which the overall line-shift is a free parameter.
539:
540:
541:
542: \subsection*{Other GRBs with X-ray spectral ``features''}
543: \noindent
544: There are GRBs of known redshift in whose X-ray data the
545: observation of ``$\rm Fe$'' lines or recombination edges has
546: been claimed; in chronological order: GRB 970508 (Piro et al., 1999), GRB
547: 970828 (Yoshida et al. 1999, 2001 and references therein) and GRB
548: 991216 (Ballantyne et al. 2002 and references therein). The corresponding
549: data are shown in Figs.~(\ref{508lines}, \ref{828lines}, \ref{216lines}),
550: from which we have, as for GRB 011211 in Fig.~(\ref{211lines}), eliminated
551: theoretical lines that unavoidably ``guide the eye''. It is clear from these
552: figures, without further ado, that
553: without a very good knowledge of the shape and magnitude of the
554: smooth continuum underlying the putative lines, it is not possible
555: to claim the observation of statistically convincing effects.
556: The evidence for a line is more convincing in the case of
557: GRB 000214 (Antonelli et al. 2000), but its redshift is not known, precluding an
558: explicit analysis.
559:
560: The case for the observation of line features is presumably weakest
561: for GRB 970508, see Fig.~(\ref{508lines}). The upper panel is
562: the spectrum of the ``early'' X-ray AG, extending for some 30 ks after the
563: start of the observations, at 6 hours after the GRB. The lower figure
564: shows the later data around 1 day after the burst. The feature at
565: $\rm E\sim 3.4$ keV in the upper panel has been interpreted as
566: an $\rm Fe$ line at the GRB's redshift $\rm z=0.835$ (Piro et al.
567: 2000). In Dado et al. (2002a) we have fit the
568: optical AG of this GRB in the CB model, the resulting
569: parameters are $\theta=3.51$ mrad, $\gamma_0=1123$ and $\rm x_\infty=0.293$
570: Mpc\footnote{The value of $\rm x_\infty$ is for the early part of the AG,
571: the time at which the X-ray
572: line was possibly observed, which precedes the abrupt
573: rise in this AG at $\rm\sim 1$ day, discussed in Dado et al. (2002a).}.
574: With these parameters, we obtain $\rm\delta(t_{obs})\simeq 142$
575: and $\rm B(t_{obs})=78$,
576: nearly constant through the X-ray observation time, as in
577: Fig.~(\ref{linemigration}). Boosted by this $\rm B(t_{obs})$, one of the
578: potentially strong lines, the $\rm n=2$
579: to $\rm n=1$ transition in $\rm HeII$ would be at 3.17 keV,
580: where the feature is in the upper panel of Fig.~(\ref{508lines}).
581: For this particular GRB, which is viewed at a relatively large angle,
582: $\delta$ and $\rm B$ are quite small, and other putative lines are at
583: sub-keV energies, where absorption appears to be very
584: strong. The data, however, are not good enough to extract conclusions from
585: the coincidence of the observed feature and the $\rm He$ line, nor from its
586: $\rm Fe$-line interpretation.
587:
588: GRB 970828, in spite of its being well localized (Remillard et al. 1997,
589: Smith et al. 1997, Marshall et al. 1997, Murakami et al. 1997, Greiner et al.
590: 1997), had no detectable optical AG down to a magnitude $\rm R\simeq 23.8$
591: (Groot et al. 1998). Such ``orphan'' GRBs are expected in the CB
592: model, not only because of possible absorption, but because
593: the time at which the optical AGs begin to decline very fast is
594: extremely sensitive to the circumburst ISM density,
595: and may be as short as ${\cal{O}}(10^{-2})$ days, see Fig.~(6) of Dado et
596: al. (2001). In that article, lacking optical data, we fit the X-ray light curve
597: of this GRB in the CB model,
598: with the result that its parameters were $\gamma_0=1153$,
599: $\theta=0.86$ mrad and $\rm x_\infty=0.87$ Mpc
600: (these values are not as well determined as in GRBs with
601: well measured optical AGs). Yoshida et al. (1999, 2001)
602: analized the spectra of this GRB in three time intervals, in the
603: middle one of which, at $\rm t_{obs}\sim 1.2\times 10^5$ s,
604: they found hints of structure, reproduced in Fig.~(\ref{828lines}).
605: These authors first ascribed the feature at $\rm E\sim 4.8$ keV
606: to the $\rm Fe$ $\rm K_\alpha$ line, which resulted in a prediction
607: of a redshift $\rm z=0.33$. The subsequently measured redshift
608: of the likely host galaxy is $\rm z=0.9578$ (Djorgosvski et al. 2001).
609: More recently Yoshida et al. (2001) attribute the feature to
610: a recombination edge of $\rm Fe$. The CB model fit results in
611: $\rm \delta(t_{obs})\sim 1004$
612: for which, at the GRBs redshift, $\rm B(t_{obs})\sim 513$, and
613: a $\rm Ly\alpha$ line would be at a predicted $\rm E\sim 5.2$ keV which is,
614: as shown in Fig.~(\ref{828lines}), quite compatible with the position of
615: the apparent feature in the data.
616:
617: In the case of GRB 991216, Piro et al. (2000) have interpreted the features
618: at $\sim 3.4$ keV and $>\! 5$ keV
619: of the X-ray spectra shown in Fig.~(\ref{216lines}) as the 6.7 keV
620: $\rm K_\alpha$ line of He-like $\rm Fe$, and a
621: $\rm Fe$ recombination edge, respectively. The reported significance
622: of the line is larger than $4\,\sigma$. On the other hand, Ballantyne et al. (2002)
623: have analized in detail the line feature in a more specific
624: model. They report that the ``F-test'' significance of the $\rm K_\alpha$ line
625: is 98\%, an explicit example of how difficult it is to convince oneself
626: that lines have actually been observed: a $\rm 2.33\,\sigma$
627: effect in a Gaussian distribution has the same significance\footnote{These
628: authors also find that an extra non-Galactic absorption ``is significant
629: {\it only} at the 97\% confidence level'' (the emphasis is ours, and 97\%
630: is $\rm 2.17\,\sigma$).}. For this GRB, the CB model fit to the optical AG
631: results in $\gamma_0=906$, $\theta=0.43$ mrad and $\rm x_\infty= 0.462$
632: Mpc, which imply, at the average observational time
633: $\rm t_ {obs}\sim 39$ hours, $\rm\delta(t_ {obs})\sim 905$
634: and a boost $\rm B(t_{obs})\sim 448$ at the
635: GRB's redshift $\rm z=1.02$. For that predicted boost,
636: there are no indications of deviations
637: from a smooth distribution at the positions of the $\rm H$ Balmer lines,
638: except, perhaps, for the corresponding recombination edge. But the $\rm n=3$
639: to $\rm n=2$ line of $\rm HeII$ and the $\rm H$ $\rm Ly\alpha$
640: line very snuggly coincide with the two allegedly significant
641: features of the X-ray spectrum,
642: as shown in Fig.~(\ref{216lines}) (the alleged Fe line centers at 3.4 keV, the
643: predicted He line is at 3.39 keV). But,
644: once again, the data are not precise enough to extract decisive inferences.
645:
646: \section*{A rough way to search for lines}
647:
648: The procedure we have discussed to predict
649: the Doppler factor, $\rm \delta(t)$,
650: is elaborate: it involves fitting the available data with use of
651: Eqs.~(\ref{sum}) to (\ref{cubic}). There is an approximate and much
652: simpler procedure that observers looking at a particular X-ray
653: AG may find useful.
654:
655: The observed X-ray frequencies are, in the
656: CBs' rest system, always above the injection bend:
657: $\rm \nu \! \gg \! \nu_b$ in Eq.~(\ref{sync}). This implies that the
658: observed energy flux of Eq.~(\ref{Fnuobser}) is:
659: \begin{equation}
660: \rm F_{obs}\propto \gamma^2\,\delta^3\,
661: \left[\nu \over \delta\right]^{-p/2}\, \nu_b^{(p-2)/2}
662: \propto \gamma^{3p/2-1}\,\delta^{3+p/2}\;\nu^{-p/2}
663: \label{approxflux}
664: \end{equation}
665: The index $\rm p$ can be fit to the X-ray spectrum of the AG
666: being studied. If this X-ray AG, as it usually the case, is observed
667: late enough for its light-curve to be an approximate
668: power law $\rm F_{obs}\propto t^{-\alpha}$, $\alpha$ can
669: be fit to the data. For this approximation to be good,
670: it is necessary that
671: $\rm \delta \approx 2\gamma$, that is $\theta^2\gamma^2\! \ll \! 1$
672: in Eq.~(\ref{doppler}), implying that Eq.~(\ref{approxflux})
673: further simplifies to $\rm F_{obs}\propto \delta^{2p+2}$.
674: The conclusion is that $\rm \delta(t) \propto t^{-b}$ with
675: $\rm b=\alpha/[2p+2]$ (for the theoretical $\rm p=2.2$,
676: $\rm \delta\propto b^{-\alpha/6.4}$, so that $\delta$ evolves roughly
677: as the 6.4-th root of the X-ray AG energy flux). This result
678: is approximate also in that Eq.~(\ref{approxflux}) describes
679: the synchrotron-radiated X-ray background, not the extra line
680: contribution. But if the latter is quite significant, the lines are
681: easier to find!
682:
683:
684: Knowing the time dependence of $\rm\delta(t)$ allows the observer
685: to ``stack'' the data taken at different times, in a search for lines
686: with an energy evolving with time as in Eq.(\ref{Lineenergy}), that is
687: $\rm E_i(t)\propto \delta(t)$. The trick is to construct a
688: time-integrated spectrum:
689: \begin{eqnarray}
690: \rm
691: \widetilde F(\widetilde \nu) & = & \rm
692: \sum_{t_{obs}}\;F_{obs}(\widetilde \nu,t_{obs})\, ,\nonumber\\
693: \rm
694: \widetilde \nu
695: & \equiv & \rm \nu_{obs}\;\left[{t_{obs}\over t_0}\right]^b\; ,
696: \label{trick}
697: \end{eqnarray}
698: where $\rm t_0$ is an arbitrary reference time, e.g. the onset of the
699: observations: $\rm t_0=min[t_{obs}]$. In $\rm \widetilde F$ the CB-model's
700: lines occur at approximately fixed, time-independent scaled frequencies
701: $\widetilde \nu$, so that the lines ``stack up'' ---rather than drifting--- with time.
702:
703:
704: \subsection*{Line intensities}
705: \noindent
706: A detailed modelling of the line intensities is a very involved
707: problem. Here we can only offer a qualitative, order-of-magnitude
708: discussion of the subject.
709:
710: The recombination rate in a hydrogenic CB is (Osterbrock 1989):
711: \begin{equation}
712: \rm R_{rec}\simeq 6.0\times 10^{44}\, x^2\,
713: \left[{N_{_{CB}}\over 6\times 10^{50}}\right]\,
714: \left[{n_b\over 10^7\, cm^{-3}}\right]\,
715: \left[{T\over 10^4\,K}\right]^{-0.7}\, s^{-1}\,,
716: \label{Recombination}
717: \end{equation}
718: where $\rm N_{_{CB}}$ is the total baryon number of the CB,
719: $\rm n_b$ is its baryon density, and $\rm x$ is the fraction
720: of ionized hydrogen in the CB.
721: The line emission luminosity in the CB rest frame is
722: $\rm R_{rec}\times 13.6\, eV\, .$
723: The corresponding radiation is boosted and relativistically beamed to an
724: observed energy flux:
725: \begin{eqnarray}
726: \rm F_{lines} &\simeq & \rm 2.5 \times 10^{-11}\;(1+z)\; erg\, cm^{-2}\, s^{-1}
727: \nonumber \\ &\times&\rm n_{_{CB}}\,
728: x^2\,
729: \left[{D_L\over 2\times 10^{28}\, cm }\right]^{-2}
730: \left[{N_{_{CB}}\over 6\times 10^{50}}\right]
731: \left[{n_b\over 10^7\, cm^{-3}}\right]
732: \left[{T\over 10^4\,K}\right]^{-0.7}
733: \left[{\delta\over 10^3}\right]^{4},
734: \label{Recombinationflux}
735: \end{eqnarray}
736: where $\rm n_{_{CB}}$ is the number of CBs (or prominent GRB pulses),
737: and the luminosity distance $\rm D_L$ is normalized
738: to its reference value
739: for $\rm z=1$ and our chosen cosmological parameters.
740:
741: For the reference parameters to which we have normalized
742: Eq.~(\ref{Recombinationflux}), the flux
743: is comparable to those reported
744: for the X-ray line-emissions in GRB afterglows. Its exact value
745: depends rather weakly on temperature and quite
746: strongly on the ionization fraction $\rm x$ in the CBs,
747: whose qualitative evolution can be assessed as follows.
748: The bound-free cross section for photoionization of atomic hydrogen in its
749: $\rm n$-th excited state by photons with frequency above the ionization
750: threshold, $\rm \nu_n=3.29\times 10^{15}/n^2\, Hz$, is given by
751: $\rm \sigma_\nu(n)=n\,\sigma_1\,\bar{g}_n (\nu/\nu_n)^{-3}$, with
752: $\rm \sigma_1= {64\, \alpha\, \pi\, a_0^2/(3\, \sqrt{3}})$ $\simeq
753: 7.91\times 10^{-18}$ cm$^2$ ($\rm a_0=0.53\times 10^{-8}$ cm
754: is the Bohr radius and $\rm \bar{g}_n$ is the Gaunt factor
755: for photo-absorption by hydrogen). Thus, a partially ionized CB
756: with a typical radius $\rm R_{_{CB}}\simeq 2.5\times 10^{14}\, cm,$ and
757: density $\rm n_b\sim 10^7\, cm^{-3},$ is opaque to optical
758: radiation. The recombination
759: photons are repeatedly reabsorbed and reemitted while diffusing out of
760: the CB. The optical radiation of a CB (X-rays in the observer's frame) is
761: the sum of the line emission and the power-law synchrotron radiation from its
762: surface. In a quasi-equilibrium
763: state the ionization fraction is such as to keep the local recombination
764: rate equal to the joint ionization rate by the synchrotron
765: and recombination radiations\footnote{
766: The cooling rate of electrons via bremsstrahlung,
767: $\rm L_{brem}\simeq 1.43\times 10^{-27}\, \bar n_e\, T^{1/2}\,
768: erg\, s^{-1}, $ is more than three orders of magnitude smaller
769: than the electron cooling rate via recombination and line emission.}.
770: The temperature is controlled by the same equilibrium, by
771: the CBs' surface energy loss and by the energy input from the continuing
772: collision of the CB and the ISM. It is difficult to ascertain without
773: a complete modelling of the problem. In Eq.~(\ref{Recombinationflux})
774: we have used a reference $\rm T$ so that the
775: maximum of the thermal distribution (at $\rm\sim\! 3\,T$) is
776: of the order of magnitude of the line energies of Hydrogen,
777: whose transitions dominate the thermal energy transport within the CB.
778:
779: Initially, the ionization is close to maximal and the line radiation of
780: Eq.~(\ref{Recombinationflux}) results in a flux comparable to that of the power-law-behaved synchrotron radiation in the X-ray band.
781: Later, when $\rm \gamma(t)$
782: decreases, equilibrium between the ionization and recombination rates
783: results in a rapid decline of line emission:
784: the recombination rate is $\rm\!\propto\!x^2$ and the photo-ionization
785: rate is $\rm\propto\! (1-x)\,\gamma^2$, so that when $\rm x$ is small,
786: $\rm x\!\propto\!\gamma$, and the recombination rate decreases like
787: $\gamma^2$. Moreover, the diffusion
788: time of recombination photons becomes very long as $\rm x\to 0$
789: which results in strong suppression of line emission.
790: The derivation of an exact X-ray spectrum and its time dependence would
791: require very complicated radiation-transport calculations which are beyond
792: the scope of this paper.
793:
794: \section*{Conclusions}
795: \noindent
796: We have studied an alternative (Dar and De R\'ujula 2001a)
797: to the interpretation by Reeves et al. (2002a)
798: of the X-ray data of XMM-Newton on GRB 011211. Unlike the quoted authors,
799: we have not for the moment studied in detail the very involved question of
800: the predicted absolute and relative intensities of the lines, which is very
801: model dependent (the density profile, ionization level and temperature
802: of a CB, as well as their time dependence, are quite complicated issues).
803: But we have shown that, in the CB model, the positions of the lines
804: are predictable and happen to coincide with the meager
805: evidence for most of them.
806: In the CB model
807: long-duration GRBs are associated with SNe that are compatible with
808: an approximately SN1998bw-like standard candle. Reeves
809: et al. (2002a) adduce that their data
810: supports this association; we contend that ---if it does--- it is not for the
811: reasons they advance, but because their observations are consistent
812: with CB-model expectations.
813:
814: We have shown that the Fe-line candidates observed in three other
815: GRBs could very well be $\rm H$ or $\rm He$ lines, again predictably
816: boosted by the very fast motion of cannonballs. The individual data
817: on each of the four GRBs that we have discussed is inconclusive, but the
818: overall consistency of the CB-model interpretation of their X-ray
819: spectral features is encouraging. In the CB model the presence of X-ray
820: lines which ---case by case--- have predictable energies,
821: is a very natural possibility. In contrast, in the
822: other scenarios that have been discussed, the X-ray lines require
823: in every instance the introduction of ad-hoc and sometimes rather exotic
824: hypothesis on the surroundings of the GRB engine.
825:
826:
827: With better data it ought to be possible to distinguish the lines expected
828: in the CB model from the ones of the standard GRB paradigm(s). Not only
829: the line positions can, in the CB model, be foretold; but also their widths
830: should be narrow, and predictably time dependent (Dar end De R\'ujula 2001a).
831:
832: \acknowledgements{We thank Ehud Behar and Ari Laor for useful
833: comments. The support of the Asher Fund for Space Research at the Technion
834: is gratefully acknowledged.}
835:
836:
837: \begin{thebibliography}{}
838: \bibitem{}
839: Antonelli, L.A., et al. 2000, ApJ, 545, L39
840: \bibitem{}
841: Ballantyne, D.R., et al. 2002, astro-ph/0206116
842: \bibitem{}
843: Belloni, T., et al. 1997, ApJ, 479, 145
844: \bibitem{}
845: Bhargavi, S.G., et al. 2001, GCN Circ. 1202
846: \bibitem{}
847: Bloom, J.S., et al. 2001, GCN Circ. 1193
848: \bibitem{}
849: Bottcher, M. 2000, ApJ, 539, 102
850: \bibitem{}
851: Bottcher, M. \& Fryer C.L. 2001, ApJ, 547, 338
852: \bibitem{}
853: Burud, I., et al. 2001, GCN Circ. 1213
854: \bibitem{}
855: Covino, S,. et al. 2002, astro-ph/0207335
856: \bibitem{}
857: Dado, S., Dar, A. \& De R\'ujula, A. 2002a, A\&A, 388, 1079
858: \bibitem{}
859: Dado, S., Dar, A. \& De R\'ujula, A. 2002b, ApJ, 572, L143
860: \bibitem{}
861: Dado, S., Dar, A. \& De R\'ujula, A. 2002c, astro-ph/0203315
862: \bibitem{}
863: Dado, S., Dar, A. \& De R\'ujula, A. 2002d, submitted to A\&A,
864: (astro-ph/0204474)
865: \bibitem{}
866: Dar, A. 1999, A\&A, 138S, 505
867: \bibitem {}
868: Dar, A. \& De R\'ujula, A. 2000a, astro-ph/0008474
869: \bibitem {}
870: Dar, A. \& De R\'ujula, A. 2000b, astro-ph/0012227
871: \bibitem {}
872: Dar, A. \& De R\'ujula, A. 2001a, astro-ph/0102115
873: \bibitem {}
874: Dar, A. \& De R\'ujula, A. 2001b, astro-ph/0105094
875: \bibitem{}
876: Djorgosvski, S.G., et al. 2001, astro-ph/0107539, submitted to ApJ
877: \bibitem{}
878: Eikenberry, S.S., et al. 2001, astro-ph/0107296
879: \bibitem{}
880: Fiore, F., et al. 2001, GCN Circ. 1203
881: \bibitem{}
882: Fruchter, A., et al. 2001, GCN Circ. 1200
883: \bibitem{}
884: Galama, T.J., et al. 1998, Nature, 395, 670
885: \bibitem{}
886: Gandolfi, G. et al. 2001, GCN Circ. 1188
887: \bibitem{}
888: Ghisellini, G., et al. 2002, A\&A, 389, L33
889: \bibitem{}
890: Gladders, M. et al. 2001, GCN Circ. 1209
891: \bibitem{}
892: Greiner, J., et al. 1997, IAU Circ. 6757
893: \bibitem{}
894: Grav, T., et al. 2001, GCN Circ. 1191
895: \bibitem{}
896: Groot, P.J., et al. 1998, ApJ, 493, L27
897: \bibitem{}
898: Henden, A.R., et al. 2002, GCN Circ. 1303
899: \bibitem{}
900: Holland, S.T., et al. 2002, Astro-ph/0202309
901: \bibitem{}
902: Hubbard, R. 2002 {\it Proceedings of the Fourth Microquasar Workshop},
903: Carg\`ese, Corsica, 2002
904: \bibitem{}
905: Jensen, B.L., et al. 2001, GCN Circ. 1195
906: \bibitem{}
907: Kotani, T., et al. 1996, PASPJ, 48, 619
908: \bibitem{}
909: Kumar, P. \& Narayan, R. 2002, astro-ph/0205488
910: \bibitem{}
911: Laor, A., et al. 1997, ApJ, 489, 656
912: \bibitem{}
913: Lazzati, D., Campana, S. \& Ghisellini G. 1999, MNRAS, 304, L31
914: \bibitem{}
915: Lazzati, D., Ramirez-Ruiz, E. \& Rees, M.J. 2002, ApJ, 572, L57
916: \bibitem{}
917: Margon, B.A. 1984, ARA\&A, 22, 507
918: \bibitem{}
919: Marshall, F.E., et al. 1997, IAU Circ. 6727
920: \bibitem{}
921: Marscher, A.P. 2002, Nature, 417, 625
922: \bibitem{}
923: Meszaros, P. \& Rees, M.J. 2001, ApJ, 556, L37
924: \bibitem{}
925: Mirabel, I.F. \& Rodriguez, L.F. 1994, Nature, 371, 46
926: \bibitem{}
927: Mirabel, I.F. \& Rodriguez, L.F. 1999, ARA\&A, 37, 409
928: \bibitem{}
929: Murakami, T., et al. 1997, IAU Circ. 6732
930: \bibitem{}
931: Osterbrock, D.E. 1989, {\it Astrophysics of Gaseous Nebulae and Active
932: Galactic Nuclei} (University Science Books, Mill Valley, Calif.)
933: \bibitem{}
934: Particle Data Group. 2000, Review of Particle
935: Physics (The European Physical Journal, 15 (2000) Numbers 1-4; Page
936: 10, Paragraph 4-2-2
937: \bibitem{}
938: Peebles, P.J.E.
939: 1993, {\it Principles of Physical Cosmology} (Princeton Univ. Press)
940: \bibitem{}
941: Piro, L., et al. 1998, A\&A, 331, L41
942: \bibitem{}
943: Piro, L., et al. 2000, Science, 290, 955
944: \bibitem{}
945: Rees, M.J. \& Meszaros, P. 2000, ApJ, 545, L73
946: \bibitem{}
947: Reeves, J.N., et al. 2002a, Nature, 416, 512
948: \bibitem{}
949: Reeves, J.N., et al. 2002b, astro-ph/0206480
950: \bibitem{}
951: Remillard, R., et al. 117, IAU Circ. 6726
952: \bibitem{}
953: Rodriguez, L.F. \& Mirabel, I.F. 1999, ApJ, 511, 398
954: \bibitem{}
955: Rutledge, R. E. \& Sako, M. 2002, astro-ph/0206073
956: \bibitem{}
957: Smith, D., et al. 1997, IAU Circ. 6728
958: \bibitem{}
959: Taylor, G.J., et al. 1997, Nature, 389, 263
960: \bibitem{}
961: Vietri, M., et al. 2001, ApJ, 550, L43
962: \bibitem{}
963: Wang, W., Zhao Y. \& You J.H. 2002, astro-ph/0203222
964: \bibitem{}
965: Weth, C., et al., 2000, ApJ, 534, 581
966: \bibitem{}
967: Yoshida, A., et al. 1999, A\&A, 138S, 433
968: \bibitem{}
969: Yoshida, A. et al. 2001, ApJ, 557, L27
970:
971:
972: \end{thebibliography}
973:
974:
975: \newpage
976:
977: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
978: \begin{figure}[]
979: \hskip 2truecm
980: \vspace*{0.3cm}
981: \hspace*{-1.6cm}
982: %\epsfig{file=fig1.eps,width=12cm}
983: \plotone{f01.eps}
984: \figcaption{Comparison between the observations in the I, R, V and
985: B bands of the optical afterglow of GRB 011211 and the CB model
986: fit as given by Eqs.~(\ref{sync}) to (\ref{cubic}). The figure
987: shows (from top to bottom) 10 times the I-band, the R-band and 1/10
988: of the V-band and 1/100 of the B-band. The line labelled X is the
989: predicted synchrotron contribution to the X-ray light curve. The
990: data are from Bhargavi et al. (2001) Burud et al. (2001), Covino
991: et al. (2001), Fiore et al. (2001), Grav et al. (2001), Jensen et
992: al. (2001), Holland et al. (2002), and Fruchter et al. (2001),
993: recalibrated with the observations of Henden et al. (2002).
994: \label{ag011211}}
995: \end{figure}
996:
997: \begin{figure}[]
998: \hskip 2truecm
999: %\vspace*{1cm}
1000: \hspace*{-1.6cm}
1001: %\epsfig{file=Borozdinus.eps,width=12cm}
1002: %\plotone{Borozdinus.eps}
1003: \plotone{f02.eps}
1004: \figcaption{X-ray spectrum of GRB 011211, as analized by Borodzin and
1005: Trudolyubov (2002). The data are from various detectors on board
1006: XXM-Newton: MOS1 (circles), MOS2 (squares) and PN (no added symbols).
1007: A power-law is only modified by absorption in the Galaxy. The dashed line
1008: is for a best-fit spectral index of 2.14, the continuous line is for the value
1009: 2.1 characteristic of the CB model.
1010: \label{211spectrum}}
1011: \end{figure}
1012:
1013:
1014: \begin{figure}[]
1015: \hskip 2truecm
1016: \vspace*{.5cm}
1017: \hspace*{-0.6cm}
1018: \epsfig{file=f03.eps,width=14cm}
1019: %\plotone{ReevesC.eps}
1020: %\plotone{f3.eps}
1021: \figcaption{(a) X-ray spectrum of GRB 011211 during the 5 ks of observations
1022: in which putative line features were observed (Reeves et al. 2002).
1023: (b) The same data with a ``background'' line scaled from a fit to the full
1024: 27 ks of observations. The vertical lines are at the predicted positions
1025: of lines in the CB model.
1026: \label{211lines}}
1027: \end{figure}
1028:
1029: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1030: \begin{figure}[]
1031: \hskip 2truecm
1032: \vspace*{.5cm}
1033: \hspace*{-1.6cm}
1034: %\plotone{ag51no1211.eps}
1035: \epsfig{file=f04.eps,width=14cm}
1036: %\plotone{lineevol.eps}
1037: %\plotone{f4.eps}
1038: \figcaption{Comparison between the predicted energies
1039: of the expected prominent lines in the X-ray afterglow
1040: of GRB 011211 as function of time after burst,
1041: as given by Eq.~(\ref{Lineenergy}), with the time-dependent Doppler
1042: factor obtained from our CB-model fit to its optical afterglow.
1043: \label{linemigration}}
1044: \end{figure}
1045:
1046: \begin{figure}[]
1047: \vspace*{.5cm}
1048: \hspace*{1.6cm}
1049: %\epsfig{file=970508us.ps,width=14cm}
1050: \epsfig{file=f05.ps,width=12cm}
1051: %\plotone{970508us.ps}
1052: \caption{The X-ray spectrum of GRB 970508. (a) In the first and (b)
1053: in the second part of the observations (Piro et al. 1998). The vertical
1054: line is at the position predicted in the CB model for the
1055: $\rm Ly\alpha$-like transition in $\rm HeII$.
1056: \label{508lines}}
1057: \end{figure}
1058:
1059: \begin{figure}[]
1060: \hskip 2truecm
1061: %\vspace*{.5cm}
1062: \hspace*{-0.5cm}
1063: %\epsfig{file=970828us.eps,width=12cm}
1064: %\plotone{970828us.eps}
1065: \plotone{f06.eps}
1066: \figcaption{The X-ray spectrum of GRB 970828 in the intermediate time-period
1067: in which a putative line feature was observed (Yoshida et al. 2001).
1068: The vertical line is at the position predicted in the CB-model for the
1069: $\rm Ly\alpha$ transition.
1070: \label{828lines}}
1071: \end{figure}
1072:
1073: \begin{figure}[]
1074: \vskip 1cm
1075: \hskip 2truecm
1076: %\vspace*{2cm}
1077: \hspace*{-0.5cm}
1078: %\epsfig{file=991216us.eps,width=12cm}
1079: %\plotone{991216us.eps}
1080: \plotone{f07.eps}
1081: \figcaption{The X-ray spectrum of GRB 991216. The top (bottom) spectrum is
1082: that of the ACIS-S (HETG) counter of
1083: {\it Chandra} (Piro et al. 1999). The vertical lines are at the positions
1084: predicted in the CB model (the $\rm He$ line is the $\rm H\alpha$-like
1085: transition in $\rm HeII$).
1086: \label{216lines}}
1087: \end{figure}
1088:
1089:
1090: \end{document}
1091:
1092:
1093:
1094: