1: \documentclass{mn2e}

2: %\documentclass[referee]{mn2e}

3: 

4: \usepackage{psfig}

5: 

6: 

7: \def\ltsima{$\; \buildrel < \over \sim \;$}

8: \def\lsim{\lower.5ex\hbox{\ltsima}}

9: \def\gtsima{$\; \buildrel > \over \sim \;$}

10: \def\gsim{\lower.5ex\hbox{\gtsima}}

11: \def\mes{M\'esz\'aros}

12: \def\gray{$\gamma$-ray}

13: 

14: \begin{document}

15: \title[Long lasting engines in GRBs]{X-ray flares and the duration of

16: engine activity in gamma-ray bursts}

17: 

18: \author[Lazzati \& Perna] {Davide Lazzati and Rosalba Perna \\ JILA,

19: University of Colorado, Boulder, CO 80309-0440, USA {\tt e-mail:

20: lazzati@colorado.edu, rosalba@jilau1.colorado.edu}}

21: 

22: \maketitle

23: 

24: \begin{abstract}

25: The detection of bright X-ray flares superimposed on the regular

26: afterglow decay in {\em Swift} gamma-ray bursts has triggered theoretical

27: speculations on their origin. We study the temporal properties of

28: flares due to internal dissipation and external shock mechanisms. We

29: first show that at least a sizable fraction of the flares cannot be

30: related to external shock mechanisms, since external shock flares

31: evolve on much longer time scales than observed. We then study flares

32: from internal dissipation, showing that the temporal properties allow

33: us to distinguish the emission of slow early shells from that of late

34: faster shells. We show that, due to the rapid evolution of the

35: detected flares, it is most likely that the flares are produced by

36: relatively fast shells ejected by the central engine shortly before

37: they are observed. This implies that the central engine must be active

38: for, in some cases, as long as one day.  We finally discuss the

39: constraints and implications that this observation has on the

40: properties and physics of the inner engine, and we elaborate on

41: possible future observational tests on the flare sample to further

42: understand their origin and physics.

43: \end{abstract}

44: \begin{keywords}

45: gamma-ray: bursts ---  radiation mechanisms: non-thermal

46: \end{keywords}

47: 

48: \section{Introduction}

49: 

50: The external shock model predicts afterglows characterized by a

51: power-law decay in time, with the possible presence of breaks

52: connecting branches of different slope (\mes~\& Rees 1997). Such

53: changes are due to either geometrical properties of the fireball

54: (geometrical beaming, Rhoads 1999) or to spectral transitions

55: (e.g. from a cooling to a non-cooling electron population, Sari, Piran

56: \& Narayan 1998). In the pre-{\em{Swift}} era, most afterglows were

57: consistent with the simplest version of this model. There were,

58: however, notable exceptions, such as GRB~000301C (Masetti et

59: al. 2000), GRB~021004 (Lazzati et al. 2002), and GRB~030329 (Matheson

60: et al. 2003). All these bursts displayed optical variability in the

61: form of bumps or wiggles superimposed on the smooth power-law decay.

62: 

63: A number of explanations have been discussed in the literature to

64: account for the variations in the optical afterglow brightness. These

65: include inhomogeneities in the external density (Wang \& Loeb 2000;

66: Lazzati et al. 2002; Heyl \& Perna 2003), refreshed shocks due to the

67: collision of a late shell of plasma with the external shock material

68: (Rees \& \mes~1998), angular inhomogeneities in the fireball energy

69: distribution (Nakar, Piran \& Granot 2002) and gravitational lensing

70: (Loeb \& Perna 1998; Garnavich, Loeb \& Stanek 2000). Different

71: mechanisms to produce variability in afterglows can in principle be

72: distinguished through their temporal and spectral properties. In

73: practice, however, a consensus has not been yet reached due to the

74: lack of unambiguous observations.

75: 

76: More recently, {\em Swift} observations have revealed that flaring

77: activity is relatively common in the early phases of GRB afterglows,

78: sometimes extending for over a day (e.g., Chincarini 2006; O'Brien et

79: al. 2006). The temporal properties of the flares, their intensity, and

80: their spectra suggest an origin that is unrelated to the external

81: shock (Chincarini 2006; Falcone et al. 2006), at least for a fraction

82: of the bumps. Late episodes of ``prompt emission''\footnote{Here and

83: in the following we will call prompt emission all the radiation

84: produced by the outflow inside the external shock. Related mechanisms

85: include, but are not limited to, internal shocks, magnetic

86: dissipation, and comptonization of external photons.} can have two

87: different origin. One possibility is that the inner engine itself is

88: active for a time as long as the detection time of the X-ray

89: flare. Alternatively, the engine can be short-lived but produce,

90: together with the fast ejecta, a tail of slower material. Such slower

91: material can produce internal dissipation (and therefore prompt

92: emission) at late time. Slow shells ejected immediately after the fast

93: ejecta will not produce an external shock at the canonical external

94: shock radius since the ambient medium has been swept by the external

95: shock produced by the fast ejecta, which develops

96: earlier. Distinguishing between these two scenarios bears important

97: implications for the physics of the GRB engine.

98: 

99: In this paper we discuss the timescales of flares in the different

100: scenarios.  We show that at least a large fraction of X-ray flares are

101: due to a long lasting activity of the GRB engine, rather than to

102: external shock activity or to the emission of slow shells immediately

103: after the prompt emission ends. 

104: 

105: This paper is organized as follows: in Sect.~2 we compute external

106: shock timescales while in Sect.~3 we consider prompt emission flares

107: and in Sect. 4 we discuss our results and compare them to {\em Swift}

108: data. We summarize our findings in Sect.~5.

109: 

110: \section{External shock}

111: 

112: Consider flaring activity produced by a sudden brightening of the

113: external shock. We consider now a brightening that involves the whole

114: surface of the external shock. Even if the duration of the activity in

115: the comoving frame is negligible, the observer at infinity still

116: detects photons over a finite amount of time, due to the ``curvature

117: effect'' (see Fig.~\ref{fig:curva}). The shape of the observed pulse

118: is therefore the narrowest possible in time. A longer activity in the

119: comoving frame will produce a longer pulse, obtained by convolution of

120: the two functional shapes.

121: 

122: \begin{figure}

123: \psfig{file=fig1.ps,width=\columnwidth}

124: \caption{{Cartoon showing the geometry of the angular time scale.}

125: \label{fig:curva}}

126: \end{figure}

127: 

128: Let us define the quantity $E^\prime(\nu^\prime)$ to be the energy

129: released by the unit area of the fireball as a result of the flaring

130: activity, in the comoving frame. The flux received at infinity is

131: given by (see also Kumar \& Panaitescu 2000):

132: %

133: \begin{equation}

134: F_{\nu}(t) = \frac{1}{4\pi\,D^2}

135: \frac{E^\prime(\nu/\delta)\,\delta^2\,d\Sigma}{dt_{\rm{obs}}}\,,

136: \label{eq:fnu}

137: \end{equation}

138: %

139: where $D$ is the distance to the source,

140: $\delta\equiv[\Gamma(1-\beta\cos\theta)]^{-1}$ is the Doppler factor,

141: $\Sigma$ is the fireball surface and $t_{\rm{obs}}$ the time in the

142: observer frame. This is given by:

143: %

144: \begin{equation}

145: t_{\rm{obs}}=\int_0^R\frac{dr}{\beta_{\rm{sh}}\,c}

146: -\frac{R}{c}\cos\theta 

147: \simeq \frac{R}{c}\left[1-\cos\theta+\frac{1}

148: {2\Gamma^2_{\rm{sh}}(2\alpha+1)}\right]\,.

149: \label{eq:tobs}

150: \end{equation}

151: 

152: Equation~(\ref{eq:tobs}) holds for any self-similar dynamical evolution

153: where $\Gamma_{\rm{sh}}\propto{R}^{-\alpha}$, as long as

154: $\Gamma_{\rm{sh}}\gg1$. Note that we indicate with the subscript

155: $_{\rm{``sh''}}$ quantities (such as Lorentz factor and speed) of the

156: external shock, and without subscript the same quantities for the

157: material just behind the shock. It can be shown that

158: $\Gamma_{\rm{sh}}=\sqrt{2}\,\Gamma$ (e.g., Sari 1997).

159: 

160: Easy considerations show that the only non constant quantity in

161: Eq.(~\ref{eq:fnu}) is the Doppler factor, since

162: $d\Sigma/dt_{\rm{obs}}=2\pi\,R\,c$, and all the radiation is emitted at

163: constant radius $R$. If we assume that we are far from a break in the

164: afterglow spectrum, $E^\prime(\nu^\prime)\propto(\nu^\prime)^{-\eta}$

165: and the functional shape of $F_{\nu}(t)$ can be rewritten as:

166: %

167: \begin{equation}

168: F_{\nu}(t)\propto\delta^{(2+\eta)}\,.

169: \end{equation}

170: 

171: Solving Eq.~(\ref{eq:tobs}) for $\cos\theta$, and expressing time in

172: unit of the start time of the flare

173: $t_0=R/[2c\,\Gamma_{\rm{sh}}^2(2\alpha+1)]$, the flare profile can be

174: written as:

175: %

176: \begin{equation}

177: F_{\nu}(t)\propto\left(1+\frac{\tau-1}{4\alpha+2}\right)^{-(\eta+2)}\,,

178: \label{eq:fnu2}

179: \end{equation}

180: %

181: where $\tau=t/t_0$. 

182: 

183: \begin{figure}

184: \psfig{file=fig2.ps,width=\columnwidth}

185: \caption{{Minimum time scale ratio for flares in different dynamical

186: settings as a function of the spectral index $\eta$. From top to

187: bottom, we show a radiative shock propagating into a uniform medium

188: ($\alpha=3$), an adiabatic shock propagating into a uniform medium

189: ($\alpha=3/2$), a radiative shock propagating into a wind profile

190: ($\alpha=1$), an adiabatic shock propagating into a wind profile

191: ($\alpha=1/2$), and the limit for internal dissipation from a slow

192: shell ejected during or immediately after the prompt phase. The shaded

193: area represents the observational distribution of $\Delta{t}/t$ from

194: Chincarini (2006), corrected for our $\Delta{t}$ definition.}

195: \label{fig:dtt}}

196: \end{figure}

197: 

198: Equation~\ref{eq:fnu2} can be used to provide a lower limit to an

199: observationally sound definition of the flare duration. Defining the

200: duration as the Full Width at Half Maximum (FWHM) of the pulse, and

201: the reference time as the time t$_{\rm pk}$ of the flare

202: maximum, we obtain:

203: %

204: \begin{equation}

205: \frac{\Delta{t}}{t} \equiv 

206: \frac{\Delta{t}_{\rm{FWHM}}}{t_{\rm{pk}}}

207: \ge \left(2^{\frac{1}{\eta+2}}-1\right)(4\alpha+2)\,.

208: \label{eq:dtt}

209: \end{equation}

210: 

211: Equation~(\ref{eq:dtt}) depends on both the dynamics and spectrum of

212: the external shock. A steeper spectrum and/or a more slowly evolving

213: fireball will cause a narrower flare. Note that in this case

214: (i.e. impulsive rebrightening of the shock), the maximum coincides

215: with the start time. Should the activity in the comoving frame last

216: for a finite amount of time, a rising phase would be observed. A more

217: detailed treatment, including the width of the shell of shocked

218: material would also provide a profile with a finite rising time.

219: Figure~\ref{fig:dtt} shows the time scale ratio for four typical

220: values of $\alpha$ as a function of the spectral index. The most

221: relevant case for X-ray flares is the adiabatic ISM case, with a

222: spectral slope $\eta\sim1.2$. We find for that specific case:

223: $\Delta{t}/t\gsim2$.

224: 

225: In an analogous context, Zhang et al. (2006) discussed the possible

226: production of flares by various mechanisms such as density bumps,

227: post-energy injection in the blast wave, and two-components or patchy

228: jets. In all cases, they concluded that the characteristics of the

229: flares were not generally consistent with what predicted by these

230: models (i.e. $\Delta t/t\sim 1$ and a slower-than observed flux

231: decline).  Here, we have derived detailed, quantitative, upper limits

232: on the value of $\Delta/t$.  Our constraints are somewhat tighter than

233: what previously derived by different authors in the context of the

234: external model for flares (e.g., Ioka, Kobayashi \& Zhang 2005) but

235: comparable to more recent results (Wu et al. 2006).  Our method

236: differs in two important ways from what done in previous work.  First,

237: we have computed the actual functional shape of the flare, and defined

238: rigorously the time reference we adopt (the peak) and the time width

239: (the FWHM). In previous works, these quantities were usually poorly

240: defined, and the results could only be approximate. Secondly, we

241: considered the difference between the Lorentz factor of the shock and

242: that of the shocked material. This factor by itself leads to flares

243: which are broader by a factor of two.

244: 

245: Chincarini (2006) presents a distribution of $\Delta{t}/t$ for a

246: sample of {\em Swift} X-ray flares that is defined as the ratio of the

247: Gaussian $\sigma$ over the peak time. The distribution is shown in

248: their Fig.~8 and shaded in Fig.~\ref{fig:dtt}. Since for a Gaussian

249: the $FWHM$ is equal to $2.35\sigma$, we can easily convert their

250: definition of $\Delta{t}$ into our definition. We conclude that the

251: distribution of $\Delta{t}/t$ of {\em Swift} X-ray flares is

252: characterized by $0.02\lsim\Delta{t}/t\lsim 1.3$ and has average

253: $\Delta{t}/t=0.3$. Considering Fig.~\ref{fig:dtt}, we conclude that

254: most, if not all, the detected rebrightenings are not related to

255: events taking place on the external shock, unless only a small portion

256: of the shock is involved in the rebrightening. Flares caused by a

257: small portion of the shock are possible, but are unlikely to be very

258: strong. Nakar \& Granot (2006) studied the effect of a blob of high

259: density interacting with the shock. They find that the rebrightening

260: is minor, and characterized by a very slow rise. Alternatively, a

261: large rebrightening on a small timescale can be observed if a narrow

262: opening angle shell refreshes a small part of the external shock

263: (Granot, Nakar \& Piran 2003). This is not likely to be the case for

264: the {\em Swift} X-ray flares, since most of them are observed before

265: the jet break time (Chincarini 2006, O'Brien 2006) and the opening

266: angle of the jet grows with time (Lazzati \& Begelman 2005; Morsony,

267: Lazzati \& Begelman 2006) rather than decrease.

268: 

269: 

270: \section{Prompt emission}

271: 

272: Since the largest majority of flares is unlikely to be produced within

273: the external shock, it must be produced during the initial phase of

274: internal dissipation. In the following we therefore discuss flaring

275: activity in the context of the prompt emission.  In this case one can

276: envisage two possible scenarios for their production: {\em (a)} shells

277: that are produced during the GRB phase but that dissipate at much

278: later times; {\em (b)} shells produced at late times by a long-lived

279: engine and dissipating on a timescale comparable to that of the engine

280: duration.  Zhang et al. (2006) also concluded that the flares have an

281: internal origin, but directly assumed that the engine must have been

282: long-lived. In this section we try to discriminate between the

283: scenarios {\em (a)} and {\em (b)}. Being able to discriminate between

284: these two scenarios bears important implications for our understanding

285: of the physics of the inner GRB engine (see more discussion on this in

286: \S4). In the following, we discuss the two scenarios above, provide

287: diagnostics for each of them, and show that the current {\em Swift}

288: data is already able to discriminate between the two flare-production

289: mechanisms.

290: 

291: {\em (a) Flaring activity due to dissipation within a

292: freely expanding flow, released during the prompt GRB phase or shortly

293: afterwards.} This case has two important differences with respect to the

294: external shock case studied above. First, there is no propagating

295: shock, and therefore there is no distinction between $\Gamma$ and

296: $\Gamma_{\rm{sh}}$. Second, the flare is produced by material that has

297: been coasting at constant $\Gamma$, rather than by material that has

298: been slowing down. These two differences change considerably the

299: equations above.

300: 

301: Without repeating the derivation of Sect.~2, it can be easily seen

302: that the result for the freely expanding flow can be obtained by

303: substituting $4\alpha+2$ with $2\alpha+1$ and adopting $\alpha=0$ in

304: Eq.~\ref{eq:dtt}.  This yields:

305: %

306: \begin{equation}

307: \frac{\Delta{t}_{\rm{FWHM}}}{t_{\rm{pk}}} \gsim

308: 2^{\frac{1}{2+\eta}}-1\,,

309: \label{eq:dtt2}

310: \end{equation}

311: %

312: which, for the typical case of $\eta\sim1.2$ yields

313: $\Delta{t}/t\gsim0.25$ (see Wu et al. 2006 for a similar computation

314: restricted to the case of internal shocks). This number is tantalizing

315: similar to the average width of {\em Swift} X-ray flares (Chincarini

316: 2006; Fig.~\ref{fig:dtt}). Yet, there is a sizable number of events

317: for which the condition is violated by almost an order of

318: magnitude. In addition, should the condition be barely satisfied, the

319: decay part of the flare would be dominated by large angle emission

320: (Kumar \& Panaitescu 2000; Liang et al. 2006). This seems not to be

321: the case, at least in some of the flares (Guetta et al. 2006). Note

322: however that the large angle emission should not dominate any flare

323: with a duration significantly larger than the one given in

324: Eq.(~\ref{eq:dtt2}).

325: 

326: {\em (b) Shell that is ejected from the central engine after a time

327: $t_{\rm{ej}}$ sizably larger than the prompt emission timescale

328: $T_{90}$.} If radiation is released at radius $R$, it will be observed

329: at a time (cfr. Eq.~\ref{eq:tobs})

330: %

331: \begin{equation}

332: t_{\rm{obs}} =

333: t_{\rm{ej}}+\frac{R}{c}\left(1-\cos\theta+\frac{1}{2\Gamma^2}\right)\,,

334: \end{equation}

335: %

336: which produces a shortening of the observed variability timescale

337: %

338: \begin{equation}

339: \frac{\Delta{t}}{t} \gsim \left(2^{\frac{1}{2+\eta}}-1\right) \,

340: \left(1-\frac{t_{\rm{ej}}}{t_{\rm{pk}}}\right)\,.

341: \label{eq:dtts}

342: \end{equation}

343: 

344: Equation~(\ref{eq:dtts}) allows for any arbitrarily small time scale, as

345: long as the shell that produces radiation is ejected at a time comparable to

346: the detection time. Therefore, the detection of fast flares implies 

347: that the inner engine is active for a time much longer than the

348: canonical prompt phase ($T_{90}$).

349: 

350: The observation of a flare that peaks at time $t_{\rm{pk}}$ and is

351: characterized by a width $\Delta{t}$ allows us to constrain the time

352: at which it was ejected from the central engine:

353: %

354: \begin{equation}

355: 1-\frac{1}{2^{\frac{1}{2+\eta}}-1}\frac{\Delta{t}}{t}

356: \lsim\frac{t_{\rm{ej}}}{t_{\rm{pk}}}\le1 \,.

357: \label{eq:wow}

358: \end{equation}

359: 

360: \section{Discussion}

361: 

362: Equations~(\ref{eq:dtt2}) and~(\ref{eq:dtts}) are general and hold for

363: any dissipation and radiation mechanism that take place in the whole

364: jet. Locally beamed emission mechanisms can produce fast variability

365: (see, e.g., Lyutikov \& Blandford 2004), but the peak luminosity of

366: the flares would drop faster than observed as a function of

367: $t_{\rm{pk}}$. Consider now the flares observed by {\em Swift} (for a

368: recent review, see {\tt

369: http://www.swift.ac.uk/rs06/Burrows.pdf}). They can be observed up to

370: $\sim10^5$~s after the GRB onset and are characterized by a FWHM time

371: $\delta{t}\sim0.3\,t$. Applying this observational constraint implies

372: that at least in about half of the cases the ejection time of the

373: material from the central engine is comparable to the time at which

374: the flare is observed. In other words, GRB engines are active well

375: beyond the observed $T_{90}$ of the prompt emission, in some cases for

376: up to $\sim10^{5}$~s or a day timescale.

377: 

378: This provides an important constraint to the properties of the inner

379: engine of long duration GRBs. More can be obtained by further analysis

380: of the properties of the flares. Flares are observed over a wide range

381: of times, spanning about three orders of magnitude (from $100$ to

382: $10^5$~s after the end of the prompt emission). The distribution of

383: $\delta{t}/t$ is however much narrower, spanning approximately only

384: one order of magnitude. This implies a correlation between the

385: ejection time and the duration of the ejection episode. The longer is

386: the time after the engine turns on, the longer is the ejection

387: episode. Interestingly, an analogous property was noted in GRBs with

388: multiple prompt emission episodes (Ramirez-Ruiz \& Merloni

389: 2001). During the prompt phase, the duration of an emission episode is

390: strongly correlated with the length of the quiescent time that

391: precedes it. In the case of flares we can conclude that the

392: correlation is with the total time since the engine onset; however,

393: this does not imply the absence of a correlation with the length of

394: the quiescent time.

395: 

396: Our results bear important implications for an understanding of the

397: physical processes governing the GRB inner engine, which provides the

398: ultimate source of energy to power GRBs. The GRB ``inner engine'' is

399: believed to be a hyperaccreting accretion disk. In the collapsar model

400: (Paczynsky 1998; MacFadyen \& Woosley 1999), this is formed by the

401: fallback material from the collapsing envelope of the massive star,

402: and the timescale for the duration of the accretion episode is set by

403: the dynamical timescale of the collapsing envelope itself, which is on

404: the order of several tens of seconds.  On the other hand, in the

405: binary merger model (e.g. neutron star-neutron star or neutron

406: star-black hole; Eichler et al. 1989), the accretion material is

407: provided by the debris of the tidally disrupted neutron star (or white

408: dwarf). In this case, the timescale over which the available material

409: is accreted is set by the viscous timescale of the disk, which is a

410: fraction of a second.

411: 

412: The accretion timescale in the collapsar model easily accounts for the

413: duration of the prompt emission for the class of long bursts, while

414: the accretion timescale in the binary merger model naturally yields

415: the required timescales for the prompt emission of short bursts.

416: Several types of observations in the last few years have indeed

417: provided a strong support for the association between long GRBs and

418: the collapse of a massive star (Stanek et al. 2003; Hjorth et

419: al. 2003), while several pieces of evidence are gradually mounting

420: toward the association of compact-object mergers and short bursts

421: (Fox et al. 2005; Bloom et al. 2006; Nakar, Gal-Yam \& Fox 2006).

422: 

423: In the scenarios above, a relativistic outflow is ejected from the

424: central engine, and radiation is produced as the bulk kinetic energy

425: is dissipated and radiated at a given radius. There is no a-priori

426: constraint on the distribution of the Lorentz factor in the outflow.

427: If the Lorentz factor of the ejecta should decrease with the ejection

428: time, late time internal activity could be seen without the need for a

429: long-living engine. We have shown that this is not the case, at least

430: in a sizable fraction of the flares.  This result makes the studies of

431: possible mechanisms of reactivation of the GRB engine especially

432: timely.  King et al. (2005) suggested that fragmentation of the

433: collapsing star could explain a (single) flare in the case of long

434: bursts, while Dai et al. (2006) proposed a new mechanism of binary

435: merger that can incorporate the presence of a flare for short bursts.

436: Perna, Armitage \& Zhang (2006) noted that the fact that flares are

437: observed in both long and short bursts, and with similar

438: characteristics in both cases, is suggestive that the place of

439: formation be in what is common between the two classes of bursts,

440: i.e. the accretion disk. A disk that fragments as a result of

441: gravitational instabilities in its outer parts, as supported by

442: studies of hyperaccreting disks (Di Matteo, Perna \& Narayan 2002;

443: Chen \& Beloborodov 2006), can account for flares with similar

444: properties in both long and short bursts, as well as accounting for

445: the observed correlation between the flare duration and its arrival

446: time (see also Piro \& Pfahl 2006).  Other disk-based models involve

447: magnetic instabilities (Proga \& Zhang 2006; Giannios 2006). Staff,

448: Ouyed \& Bagchi (2006) proposed instead a mechanism based on state

449: transitions in the quark-nova scenario, while the role of a magnetar

450: was discussed by Gao \& Fan (2006) and Cea (2006).

451: 

452: 

453: \subsection{Refreshed shocks}

454: 

455: If an ``internal'' flare, i.e., a flare due to internal dissipation,

456: is observed in a GRB, it must be followed at later time by an external

457: flare. The material responsible for the internal flare keeps expanding

458: at constant speed, and must eventually catch up with the external

459: shock, which decelerates continuously. It may however be difficult to

460: observe the second flare, since it may be weak and dispersed over

461: several orders of magnitudes in time. If, however, the second flare

462: can be singled out and firmly associated to an external flare, it

463: would provide additional important information for our understanding

464: of the dynamics of the system.

465: 

466: Consider a flare system made by an internal flare beginning at time

467: $t_{\rm{int}}$ followed by a second, broader, flare starting at time

468: $t_{\rm{ext}}$, where $t_{\rm{ext}}\gg{t}_{\rm{int}}$. The external

469: flare time can tell us the Lorentz factor of the material producing

470: the internal flare:

471: %

472: \begin{equation}

473: \Gamma_{\rm{ej}} \simeq \left(\frac{3E}{8\pi\,n\,m_p\,c^5}\right)^{1/8}

474: \,t_{\rm{ext}}^{-3/8}\,,

475: \label{eq:gamma}

476: \end{equation}

477: %

478: where $E$ is the isotropic equivalent kinetic energy of the fireball

479: integrated up to the time of ejection of the late shell, $n$ the

480: numeric density of the external medium, and $m_p$ the proton mass. We

481: have assumed a uniform external medium. Equations can be easily

482: generalized to an arbitrary density profile, but will be simple only

483: in the case of adiabatic evolution. Equation~(\ref{eq:gamma}) depends

484: only on measurable quantities.  An example of a solvable system is the

485: flare system of GRB~050502B. Falcone et al. (2006) were able to

486: identify an initial strong flare and a possible refreshed shock flare

487: at later times. They were able to measure the bulk Lorentz factor of

488: the late engine ejecta to be $\Gamma_{\rm{ej}}\lsim20$, ejected from

489: the inner engine approximately $300$ second after the onset. By itself

490: this observation only tells us that the late ejections are likely to

491: have a smaller Lorentz factor than the early ones. Solving a large

492: numbers of flare systems in different GRBs would however provide us

493: with an ensemble of late ejections that would allow us to constrain

494: both the properties of the engine and of the dissipation/radiation

495: mechanism. The knowledge of the bulk Lorentz factor of the late shells

496: would provide us with a sample of flares for which the comoving

497: spectra are known, ridding the prompt emission models of one of the

498: unknown parameters: the Lorentz factor.

499: 

500: 

501: \section{Summary}

502: 

503: We have analyzed the timescale of flares produced by several

504: mechanisms both due to internal dissipation and external shock

505: phenomena. We have shown that all prominent external shock flares

506: occurring before the jet break must have a long time scale

507: $\delta{t}\gsim{t}$. We also showed that internal dissipation taking

508: place in slow shells ejected immediately after the end of the prompt

509: phase must satisfy a similar constraint: $\delta{t}\gsim{0.25t}$. The

510: only mechanism capable of producing flares with a short time scale as

511: observed by {\em Swift} is identified with late activity of the inner

512: engine. As a consequence, the detection of X-ray flares at times as

513: long as $10^5$ seconds after the GRB onset implies that the inner

514: engine is active for at least $10^5$~s. This has important

515: implications for the fueling of the engine and the ejection mechanism.

516: 

517: \section*{Acknowledgements}

518: 

519: We thank Dafne Guetta and Luigi Piro for useful conversations.  This

520: work was supported by NASA Astrophysical Theory Grant NNG06GI06G (DL),

521: NSF grant AST-0307502 (DL) and AST 0507571 (DL \& RP), and {\em Swift}

522: Guest Investigator Program NNX06AB69G (DL) and NNG05GH55G (DL \& RP).

523: 

524: \begin{thebibliography}{}

525: \bibitem{} Bloom J. et al. 2006, ApJ, 638, 354

526: \bibitem{} Cea P., 2006, ApJ in press (astro-ph/0606086)

527: \bibitem{} Chen W.-X., Beloborodov A. M., 2006, ApJ in press,

528:   (astro-ph/0607145)

529: \bibitem{} Chincarini G., 2006, Proceedings of the Vulcano Workshop

530:   2006, Vulcano, Italy, May 22-27, Edited by F. Giovannelli \&

531:   G. Mannocchi (astro-ph/0608414)

532: \bibitem{} Dai Z. G., Wang X. Y., Wu X. F., Zhang B., 2006,

533:   Science, 311, 1127

534: \bibitem{} Di Matteo T., Perna R., Narayan R., 2002, ApJ, 579, 706

535: \bibitem{} Eichler D., Livio M., Piran T. Schramm D. N., 1989,

536:   Nature, 340, 126

537: \bibitem{} Falcone A.~D., et al., 2006, ApJ, 641, 1010

538: \bibitem{} Fox D. B., et al., 2005, Nature, 437, 845

539: \bibitem{} Gao W.-H., Fan Y.-Z., 2006, ChJAA, 6, 513

540: \bibitem{} Garnavich P.~M., Loeb A., Stanek K.~Z., 2000, ApJ, 544, L11

541: \bibitem{} Giannios D., 2006, A\&A, 455L, 5

542: \bibitem{} Granot J., Piran T., Sari R., 1999, ApJ, 513, 679

543: \bibitem{} Granot J., Nakar E., Piran T., 2003, Nature, 426, 138

544: \bibitem{} Guetta D., D'Elia V., Fiore F., Conciatore M. L., Antonelli

545:   L. A., Stella L.. 2006, Il Nuovo Cimento in press

546:   (astro-ph/0610512)

547: \bibitem{} Heyl J. S., Perna R., 2003, ApJ, 586L, 13

548: \bibitem{} Hjorth J., et al., 2003, Nature, 423, 84

549: \bibitem{} Ioka K., Kobayashi S., Zhang B., 2005, ApJ, 631, 429

550: \bibitem{} King A., O'Brien P. T., Goad M. R., Osborne J.,

551:   Olsson E., Page K., 2005, ApJ, 630L, 113 

552: \bibitem{} Kumar P., Panaitescu A., 2000, ApJ, 541, L51

553: \bibitem{} Lazzati D., Rossi E., Covino S., Ghisellini G.,

554:   Malesani D., 2002, A\&A, 396L, 5

555: \bibitem{} Lazzati D., Begelman M. C., 2005, ApJ, 629, 903

556: \bibitem{} Liang E. W., et al., 2006, ApJ, 646, 351

557: \bibitem{} Loeb A., Perna R., 1998, ApJ, 495, 597

558: \bibitem{} Lyutikov M., Blandford R., 2004, ASPC, 312, 449

559: \bibitem{} Matheson T., et al., 2003, ApJ, 599, 394

560: \bibitem{} MacFadyen A. I., Woosley S. E., 1999, ApJ, 524, 262  

561: \bibitem{} Masetti N., et al., 2000, A\&A, 359, L23

562: \bibitem{} \mes~P., Rees M.~J., 1997, ApJ, 476, 232

563: \bibitem{} Morsony B. J., Lazzati D., Begelman M. C., 2006, ApJ

564:   submitted (astro-ph/0609254)

565: \bibitem{} Nakar E., Piran T., Granot J., 2003, NewA, 8, 495

566: \bibitem{} Nakar E., Granot J., 2006, MNRAS submitted

567:   (astro-ph/0606011)

568: \bibitem{} Nakar E., Gal-Yam A., Fox D. B., 2006, ApJ, 650, 281

569: \bibitem{} O'Brien P.~T., et al., 2006, ApJ, 647, 1213

570: \bibitem{} Paczynsky B., 1998, ApJ, 494L, 45

571: \bibitem{} Perna R., Armitage P. J., Zhang B., 2006, ApJ, 636L, 29

572: \bibitem{} Piro A. L., Pfahl E., 2006, ApJ submitted (astro-ph/0610696)

573: \bibitem{} Proga D., Zhang B., 2006, MNRAS, 370L, 61

574: \bibitem{} Ramirez-Ruiz E., Merloni A., 2001, MNRAS, 320, L25

575: \bibitem{} Rees M.~J., Meszaros P., 1998, ApJ, 496, L1

576: \bibitem{} Rhoads J.~E., 1999, ApJ, 525, 737

577: \bibitem{} Sari R., 1997, ApJ, 489, L37

578: \bibitem{} Sari R., Piran T., Narayan R., 1998, ApJ, 497, L17

579: \bibitem{} Staff J., Ouyed R., Bagchi M., 2006, ApJ submitted

580:   (astro-ph/0608470)

581: \bibitem{} Stanek K., et al. 2003, ApJ, 591L, 17 

582: \bibitem{} Wang X., Loeb A., 2000, ApJ, 535, 788

583: \bibitem{} Wu X. F., Dai Z. G., Wang X. Y., Huang Y. F., Feng L. L.,

584:   Lu T., 2006, ApJ submitted (astro-ph/0512555)

585: \bibitem{} Zhang  B., et al., 2006, ApJ, 642, 370

586: \end{thebibliography}

587: 

588: \end{document}

589: