Diapositiva 1

1
High Energy Emission from Gamma-Ray Bursts
Alessandra Galli
MAGIC Gamma-Ray Burst Workshop
3 December 2007 Centro Astrofisica La Palma,
Canary Island of La Palma
2
Very High Energy emission in GRBs
There are previous observational evidences of
high energy emission in GRBs

• GRB 940217 (Hurley et al. 1994) detected by
  EGRET presents VHE emission at hundreds-thousands
  of s after GRB onset, including a 18 GeV photon

3
Very High Energy emission in GRBs

• GRB 941017 (Gonzalez et al. 2003) detected by
  EGRET presents a distinct multi-MeV spectral
  component which decays more slowly than the low
  energy component

4
Very High Energy emission in GRBs

• GRB 970417 (Atkins et al. 2000, 2003) the
  excess of events detected by Milagrito if true
  must be due to photons of at least 650 GeV.
• VHE observation are not consistent with the
  extrapolation of the BATSE spectrum. The VHE
  fluence is at least 10 times greater than the
  sub-MeV fluence observed by BATSE.

5
Very High Energy emission in GRBs

• SSC in internal shocks, 1 MeV-10 GeV (Meszaros
  et al., Galli Guetta 2007)

• SSC in RS, keV-GeV (Kobayashi et al. 2007

• SSC in FS, MeV-TeV (Galli Piro 2007)

• p-? interaction, MeV TeV (Gupta Zhang 2007)

• p-? interaction in FS, GeV TeV (Boettcher
  Dermer 2003)

6
Prompt Emission
7
Internal Shock model

• The source produces a wind approximated as a set
  of discrete shells.
• Shells are emitted with a contrast Lorentz
  factor in this model it is assumed that ?GG.
• The average interval between consecutive shell
  ejections is tv. As a consequence the flow
  Lorentz factor varies on a temporal scale tv.
• The emission properties, i.e. the peak energy,
  are determined by the radius of collisions,
  RG2ctv, and the kinetic energy of the shell.
• The peak energy Ep thus constrain the radius of
  collisions.

8
Synchrotron and inverse Compton emission from IS
(Prescriptions of Guetta Granot 2003)

• Ep, F, EpSC and Ecut strongly depend on the
  luminosity L, temporal variability tv, and
  Lorentz factor G

Ep 1.910-3(1Y)-1/3L491/2 G2-2
tv-1ee,-13/2eB,-11/2 keV
EpIC 7.510-3 L491/2 G2-2 tv-1 ee,-17/2eB,-11/2
MeV
Ecut 250 L49-9/10 G228/5 tv (ee,-13eB,-1)-1/10
GeV
(p2.5, z1)
FpIC 510-10L49 ee,-11/2eB,-1-1/2 erg cm-2s-1
prompt L1052 erg/s, G300, tv10 ms
Rcoll2G2ctv1013 cm
9
Internal absorption
XRT
BAT
Galli Guetta 2007

• Bursts with peak energy close to the lower value
  of the BAT energy band are the best candidates
  for GeV emission.
• Due to internal pair-production MAGIC cannot
  detect high energy emission during the prompt
  emission phase of GRB. High Lorentz factor
  values, G500, and tv would be required.

10
Afterglow Emission
11
ES IC versus synchrotron emission (Galli Piro
2007)
YLIC/Lsyn
FAST COOLING ?1, Y?(ee/eB)1/2
SLOW COOLING ?lt1, Y?(?ee/eB)1/2
eB (E53 n)-1/4 1(ee/eB)1/2-1/2 (ee/eB) -1/2
Td1/4 (1z)-1/4
Importance of IC emission improves for larger
(ee/eB)
12
Inverse Compton component from the afterglow
E531, n5.0, p2.5, z1
Galli Piro 2007
Tint500 sec Fth,LAT(1GeV)210-9 erg cm-2s-1
Tint104 sec Fth,LAT(1GeV)10-10 erg cm-2s-1
13
IC emission from afterglow application to GRB
940217
High energy emission, 30 MeV-30 GeV, up to 5000
sec (Hurley et al. 1994) best fit power law
?2.83?0.64 ? F 210-9 erg cm-2 s-1
E53 5.0, n 3.0, ee 0.07, eB 0.001, p
2.5, z 1
500 sec
5000 sec
F t1/3 v lt ?c,IC F t1/8 ?c,IC lt v lt
?i,IC Ft-(10-9p)/8 ?i,IC lt ?
?c,IC lt ?i,IC lt ?obs ?(p2)/2 p2.5
? ?2.25
14
Prompt-to-Afterglow Emission
15
GRB prompt-to-afterglow transition X-ray flares
Prompt-to-afterglow transition characterized by
initial steep decay, flattening, and flares.
Flare are very common, they are present in
30 -40 of the Swift GRBs sample.
X-ray flares are likely to trace the activity of
the central engine, thus they can give important
information about the physics of the progenitor.
16
X-ray Flare Properties
Temporal properties (Chincarini et al. 2007) -
flare duration 10 times shorter than the flare
peak time, i.e. flares are very rapid
phenomena - the strongest flares appear at
shorter timescales - the majority of flares
appear within 1000 sec after the burst -
several bursts present multiple flares.

• Spectral properties
• (Falcone et al. 2007)
• fitted both by a simple power law and Band law
• - Globally soft spectrum peak of the emission in
  the soft X-ray, or between the optical and the
  X-ray
• - several present strong hard-to-soft spectral
  evolution. Others have not evolving spectra
• - fluence can be as high as that brought by the
  prompt emission.

17
Possible Scenarios GROUP I

• DO NOT REQUIRE A LONG DURATION ENGINE ACTIVITY
  
• -Late Internal Shock (LIS) from a short duration
  central engine the central engine emits a tail
  of slower shells which collide at late times
  (Zhang et al. 2006, Wu et al. 2006)
• Refreshed Shock the central engine releases its
  energy with a variety of Lorentz factors. The
  faster shell is decelerated and the slower part
  of the outflow can catch up with it at later time
  injecting energy in the blast wave (Rees
  Meszaros 1998, Kumar Piran 2000, Guetta et al.
  2007)
• Forward-Reverse shock (FS-RS) for appropriate FS
  and RS shock parameters the RS can dominate in
  the X-ray (Fan Wei 2005)
• External shock on a clumpy medium external shock
  on small radii clouds can produce high variable
  GRB light curves and flares (Dermer 2007)

18
Possible Scenarios GROUP II

• LONG ACTIVITY AND/OR RE-ACTIVATION OF THE CENTRAL
  ACTIVITY
• Late Internal Shocks late internal shocks
  produce a long duration prompt emission lasting
  for hundreds to thousands of seconds (Burrows et
  al. 2005, Wu et al. 2006)
• Delayed External Shock (DES) the onset of
  afterglow emission is delayed by a long lasting
  central engine activity. The X-ray flare is
  produced in external shocks and marks the
  beginning of the afterglow (Piro et al. 2005,
  GalliPiro 2006)

19
X-ray flare studies

• - Flare spectrum softer than the main pulse and
  consistent with the afterglow spectrum at 1 day
  (Piro et al. 2005).
• -The flare connect to the afterglow with a power
  law if the origin of the time is shifted to the
  time of the flare appearance (Piro et al. 2005).

Piro et al. 2005
20
Delayed External Shock scenario thick shell
fireballs

• The onset of the external shock depends on the
  dynamical regime of the fireball, i.e. on its
  Lorentz factor G0 and on thickness ?cteng.
• In thick shells the Reverse Shock ends crossing
  the shell after the fireball starts to
  decelerate, thus teng gttdec (Sari Piran 1999).
• The crossing of the RS increases the emitting
  volume of the shell, thus the most of the energy
  is transferred to the surrounding material at
  teng (Sari Piran 1999), around the end of the
  engine activity Delayed External ShocK.
• In this case the afterglow decay is described by
  a power law only if the time is measured from the
  instant of the central engine turns off (Lazzati
  Begelman 2006).
• The flare is produced by an external shock caused
  by an energy injection lasting until the time of
  the flare occurrence, i.e. requires long lasting
  central engine activity (Galli Piro 2006)

21
Delayed External Shock application to GRB
011121 and XRF 011030

GRB 011121
XRF 011030
22
High energy emission from X-ray flares

• DES can explain only one flare. In presence of
  multiple flares additional mechanisms are
  required, e.g. late internal shocks (Burrows et
  al. 2005).
• In this case models where the flare is produced
  by the same mechanism responsible of the prompt
  emission, i.e. Late Internal Shocks, are favored.
  
• The validity of these models can be tested
  comparing their predictions at very high
  energies, from hundred of MeV to hundred of GeV
  (the energy range investigated by AGILE and
  GLAST) .
• The DES and the LIS predict different high energy
  flare properties. Simultaneous flare observations
  in the X-ray band and at high energies are thus
  very important to validate and/or discriminate
  these models.
• In both the models X-ray flare photons can be
  inverse Compton scattered by the population of
  electrons that produces the X-ray flare itself,
  and a flare arises at very high energies, MeV to
  GeV (Galli Piro 2007, AA in press Galli
  Guetta, 2007, AA submitted).

23
Predictions for X-ray and GeV flares
(Prescriptions of Panaitescu Kumar 2000)
Ep 3103 E541/2 ee,-12eB,-21/2 t-1 keV
Ecut 0.5 E53-1/4 n13/20 ee,-16/5
eB,-21/10(1Y)4/5 t1/20 TeV
(p2.5, z1)
If EobsgtEp ? F?Fp (Eobs/Ep)-p/2
(Ecol/Ep) eeee(E54,F1keV,Ep,tobs) ,
eBeB(E54,F1keV,Ep,tobs)
Observed quantities
EpIC 396 n0-1/4 E5411/12 (Ep/1keV)8/3(F1keV/1mJy)
-4/3 tobs-1/12 GeV
EpIC FpIC 1.310-7 E544/3(Ep/1keV)25/12(F1keV/1m
Jy)-5/3 (Fp/1mJy) tobs-1/12 erg cm-2s-1
24
E540.01, G100, n50, ee0.1, eB10-4, z0.1
LAT, Tint200 s
syn
IC
Ep 10 eV ? EpIC 1GeV
MAGIC
EpIC below the observational band ? strong
temporal correlation
25
E540.01, G100, n50, ee0.25, eB10-5, z1
Ep 1KeV ? EpIC 30 GeV
EpIC above the observational band ? temporal delay
26
Flares in the context of the Late Internal Shock
model
1st hypothesis flare temporal variability
measured by its duration.
Ep is in the soft X-ray, or between optical and
X-ray consistently with observations (Falcone et
al. 2007), implies G1025
The peak energy and the cutoff energy due to pair
production go below 1 GeV
BAT
NO HIGH ENERGY FLARES ARE EXPECTED
Rflare 2G2ctv 1014 cm for G10
27
2nd hypothesis flare temporal variability
similar to that of the prompt emission
Peak of the emission consistent with
observations if one assumes a Lorentz factor as
G100
The peak of IC at lower energies, tens-hundreds
of MeV, with respect to the DES
Under these assumptions X-ray flares are produced
by the many collisions, and their light curves
should present some substructures as those
observed during the prompt emission.
28
Conclusion

• Prompt emission due to internal pair production
  MAGIC is expected to be not able to detect high
  energy emission during this phase. A detection
  would be possible only very large Lorentz factor
  values, and/or tv gt 1ms.
• X-ray flares are expected to have high energy
  counterparts. In the LIS model high energy flares
  peak at lower energies (tens-hundred MeV) with
  respect to the DES model (GeV-TeV)
• In the LIS there the flare temporal variability
  and the peak energy of the IC component are
  anti-correlated. If tv is of the same order of
  magnitude of the flare duration no high energy
  flare is expected
• In the context of the DES MAGIC have good
  possibilities to detect high energy flares
• In the context of the LIS MAGIC could detect
  high energy emission from flares only for low tv

29
Bonus Slides
30
XRF 011030 in presence of Inverse Compton emission

• Model parameters are better constrained
• In presence of IC emission we can discriminate
  between can uniform interstellar medium and a
  wind like medium

Galli Piro, 2007
synch
IC
ISM case without IC E53 0.03, G0 130, n 5,
ee 0.29, eB 8 10-5, p 2.1, z 1 and Tb
8 105 sec. Y 27 X-ray decreases as (1Y),
optical and radio not affected by IC
Wind case without IC E53 0.3, G060, A0.055,
ee0.02, eB0.001, z1 and p2.1. Y 2 , fast
to slow cooling transition after the flare
31
However at lower redshift
there are good possibilities for the detection
of HE emission also by MAGIC
32
As for the DES the detectability of HE flares
improves with larger (ee/eB) values
A. Galli- Rome, 14 November 2007
33
Delayed External Shock GRB 050904
Flare spectrum consistent with the afterglow
emission. The DES model has problems to account
for the temporal gap between the prompt emission
and the flare, and to describe the rise of the
flare. In addition, the extrapolation of the
X-ray flare to the optical band below the TAROT
data no single emission component model can fit
(Boer et al. 2007). Two component models needed.
34
DES IC versus synchrotron emission (Galli Piro
2007)
YLIC/Lsyn
FAST COOLING ?1, Y?(ee/eB)1/2
SLOW COOLING ?lt1, Y?(?ee/eB)1/2
eB810-2 (E53 n)-1/4 1(ee/eB)1/2-1/2
(ee/eB) -1/2 Td1/4 (1z)-1/4
Importance of IC emission improves for larger
(ee/eB)
Detectability of IC improves for larger (ee/eB)
if the peak of emission is in the band of the
detector.
35
MODEL PARAMETERS EFFECTS
Energy E53 and fraction of energy going into
electrons ?e
36
Lorentz Factor ??
37
Density n
38
Energy fraction going into magnetic
field ??
39
High energy emission from X-ray flares
X-ray flares can be Inverse Compton scattered
producing flares in the GeV-TeV band.

• Late Internal Shock model
• 1) X-ray flares by synchrotron, and GeV flares
  by self IC emission
• ?
• strong temporal correlation
• 2) X-ray flares by synchrotron and
• GeV flares by IC on the afterglow electrons.
• ?
• No temporal correlation

Delayed External Shock model X-ray flares by
synchrotron. GeV flares by self-IC emission of
flare photons on afterglow electrons. ? strong
temporal correlation
40
XRF 011030 broad-band analysis X, optical and
radio
ISM case
Radio
X
Radio
Optical
A. Galli-Rome, 14 November 2007
41
XRF 011030 broad-band analysis X, optical and
radio
Wind case
A. Galli-Rome, 14 November 2007
42

• In 1997 BeppoSAX discovered a fading emission
  following the GRB occuring in X-ray, optical and
  radio. It remains detectable for days to weeks.
• The photons received during the classical GRB
  phenomenon are called prompt emission and the
  subsequent fading emission is called afterglow
  emission

• Prompt hard-to-soft spectral evolution
• Afterglow power law temporal decay and soft
  power law spectra

43

The FIREBALL Model
External Shock
Interstellar medium
FS
RS
1012 cm
1014 cm
1017 cm
ms time variability require a compact source, the
huge amount of energy require a plasma in
ultra-relativistic (? gt 100) expansion (Fireball)
kinetic energy converted in relativistic
electrons (ee) and magnetic fields (eB)
The progenitor loses the external layers of
material in form of shells emitted with different
relativistic velocities this causes Internal
Shocks, which produce the prompt emission
Shells interact with the external medium thus
causing an External Shock which produce the
afterglow emission. A Forward Shock and Reverse
Shock develop.