GLAST Science Lunch
Dec 1, 2005
Can emission at higher energies provide
insight into the physics of shocks and how
the GRB inner engine works?
Eduardo do Couto e Silva
Dec 1 , 2005
E. do Couto e Silva
1/21
GLAST Science Lunch
Dec 1, 2005
GRB Fundamental Questions
• Origin
– Where does the energy come from?
– How can it be so large?
– How does the inner engine work?
• Evolution
– How does the GRB evolve?
– What are the dynamics of the relativistic flow?
– How does it collimate the flow (in case of jets)?
• Observed Radiation
– What is behind the emission mechanism?
– How can we explain the radiation we observe?
E. do Couto e Silva
2/21
GLAST Science Lunch
Dec 1, 2005
Some GRB Experimental Facts
•
Intensity, location, rate
– Typical fluences
• 10-4 to -7 ergs cm-2
– Cosmological distances
• z~1
– Rate
• 1 /Myrs/galaxy
•
Energy Spectrum
– Non-thermal emission
• up to g rays (3 GeV)
•
Temporal properties
– Rapid flux variations
• miliseconds
– Range of burst durations
• few seconds to hours
E. do Couto e Silva
940217
(Hurley 1994)
Transients
are hard to catch!
3/21
GLAST Science Lunch
Dec 1, 2005
Duration of GRBs
160
(e.g stellar mass BH
accreting from a
massive disk,
rotating NS driving
Poynting flux)
120
NUMBER OF BURSTS
If there is a compact
object at the inner
engine, the source must
also be active for a long
time
Long
Compact
Mergers?
(ISM)
Collapsar
Model ?
(wind)
1 cm-3
103-4 cm-3
80
40
0
0.01
E. do Couto e Silva
Short
0.1
1
10
DURATION, SECONDS
100
1000
4/21
GLAST Science Lunch
Dec 1, 2005
Introduction to the Fireball Model
Catastrophic event:
merger or hypernova
(rapid and energetic)
Energy release within
gravitational radius of BH
(gg ↔ e+e- in equilibrium,
radiation can not escape)
 >>1
(adiabatic expansion)
 =1
gg ↔ e+ein equilibrium
~107 cm
~1013 cm
Little rest mass energy
(mostly relativistic flow)
Can it explain non-thermal
emission in g rays?
E. do Couto e Silva
5/21
GLAST Science Lunch
Dec 1, 2005
Fireball Model: quasi thermal spectrum
 gg 
f p T FD 2
(Goodman 1986)
Ri2 me c 2
Fireball
(Piran 1999)
opaque radiation plasma
whose initial energy is
significantly greater than its
rest mass
Large opacity means g
rays can’t escape !
no baryons
Tobs ~ Tinitial



F
D



 gg  1013 f p  7
2 
 10 erg cm  3000 Mpc 
E. do Couto e Silva BATSE fluences
2
 T 


10
ms


2
Burst variability
~1028 cm : cosmological distances
6/21
GLAST Science Lunch
Dec 1, 2005
GRB Fundamental Questions
• Origin
– Where does the energy come from?
– How can it be so large?
– How does the inner engine work?
• Evolution
– How does the GRB evolve?
– What are the dynamics of the relativistic flow?
– How does it collimate the flow (in case of jets)?
• Observed Radiation
– What is behind the emission mechanism?
– How can we explain the radiation we observe?
E. do Couto e Silva
7/21
GLAST Science Lunch
Dec 1, 2005
Baryonic Load in the Fireball Model
Entropy per baryon h = E0/ M0c2
High load
 >>1
 =1
(model dependent)
High: M0 > 10-9 Msun , h < 10-5
Low load
G constant
Low: M0 < 10-9 Msun, h > 10-5
Internal Energy is converted into
kinetic energy of baryons
G increases linearly with R
(Baryons limit increase in G)
Competing processes:
Bulk acceleration
relativistic flows of large G
Cooling Losses
synchrotron or IC ~ G2
~1013 cm
Internal Energy
Photons
B field
Particle pressure
neutrinos
~1010 cm
Amount of baryons is
determined by the original
environment in which the
fireball was created
E. do Couto e Silva
8/21
GLAST Science Lunch
Dec 1, 2005
Fireball Evolution
•
Expansion phase (G ~ r)
– limited by the rest mass of
baryons
•
Coasting phase (G ~ constant)
– form shells of comparable size of
the fireball at rest (Blandford and
Mckee 1976)
• Newtonian for large load
• Relativistic for small load
•
Decelerating phase (G ~ r-2 )
– limited by inertia of external
medium
Kobayashi, Piran & Sari 1999
z=43
E=1052 ergs
h=50
R =3×1010 cm
UInt to UKin
E. do Couto e Silva
Ukin to UISM
9/21
GLAST Science Lunch
Dec 1, 2005
GRB Fundamental Questions
• Origin
– Where does the energy come from?
– How can it be so large?
– How does the inner engine work?
• Evolution
– How does the GRB evolve?
– What are the dynamics of the relativistic flow?
– How does it collimate the flow (in case of jets)?
• Observed Radiation
– What is behind the emission mechanism?
– How can we explain the radiation we observe?
E. do Couto e Silva
10/21
GLAST Science Lunch
Dec 1, 2005
Dissipation Phase: Internal Shocks
•
•
Engine creates an outflow wind
– Shells are produced
• Fast ones catch slow ones
• Fluctuations in the distribution G
is responsible for time structure
•
Improve efficiency
– multiple collisions
Kobayashi & Sari 2001
Efficiency Problem
– Conversion of Ukinto radiation ~ few %
(Kumar 1999)
– can’t transfer heat into radiation from
Coulomb ep collisions
• Hydrodynamic (1/10)
• Radiation by e- (1/3)
• Observed energy band (1/3)
•
Variability depends on distance from inner
engine (Zhang & Meszaros 2003)
– Is shell’s B local or from inner engine?
E. do Couto e Silva
11/21
GLAST Science Lunch
Dec 1, 2005
Dissipation Phase: External Shocks
Kobayashi, Piran & Sari 1999
COLD
HOT
HOT
COLD
shock
ISM
HOT
COLD
RRS
Reverse Forward
shock
COLD
ISM
z=0.1
E=1052 ergs
h=104
R =4.3×109 cm
•
Plethora of models
– Shells
• thin or thick
– Shocks
• Relativistic or Newtonian
– External Medium
• ISM or wind like
E. do Couto e Silva
12/21
GLAST Science Lunch
Dec 1, 2005
What is the ambient density?
•
Deceleration Phase (~ 1017 cm)
– Stellar wind in massive progenitor
• Medium is not homogeneous
• n ~ 1/r2
• Wind sweeps ~ 1/ G x its rest
mass
• Typically ~ 10-6 MQ
– Dense ISM as in early galaxies?
• Constant medium density
• ISM now ~ 1 cm-3
• Molecular clouds > 103 cm-3
• Clumpy ISM?
•
dM
 106 MQ  yr 1
dt
vwind  103 km  s 1
SNR > 1 MQ
Soderberg and Ramiro-Ruiz 2003
a problem for
collapsar
models?
1017 cm-3
Need initial Lorentz factor to put
constraints on the swept-up matter
versus radius
SNR ~ many pc
– Timescale is days
• relativistic effect
E. do Couto e Silva
13/21
Fit parameters
GLAST Science Lunch
Dec 1, 2005
High Energy Emission Models
Leptonic
Models
(IC)
Internal
SSC
External
Forward
•
Controversial
•
Hard to model IC
Hadronic
models
(Pe’er Waxman 2005)
– KN regime
– HE EM cascades
– Large  from e+-
Proton
Synchrotron
pp, pn , pg
Reverse
•Internal
–100 keV + ge3 = 100 GeV
•External Forward
–10 keV + ge5 = 100 TeV
•External Reverse
–100 eV + ge3 = 100 MeV
E. do Couto e Silva
14/21
GLAST Science Lunch
Dec 1, 2005
Shock Plasma Parameters
•
•
•
•
•
•
Luminosity of the Source
Can we use GLAST
– L ~ 1052 ergs s-1
measurements to
• too hard for GLAST
constrain these
Lorentz Factor of Bulk Plasma
parameters?
– G ~ 102 to 103
• cut-offs at the energy spectrum (GLAST)
Time Variability
– DT ~ 10-4 to 10-3 s
• ~30 ms with GLAST
Fraction of Thermal Magnetic Energy Density behind the shock
– eB ~ 1 to 10-5
• Ratio of peaks in SED (GLAST+?)
Fraction of Thermal Electron Energy Density behind the shock
– ee ~ 1 to 10-5
Are p, eB ee time
• Ratio of peaks in SED (GLAST+?)
independent?
Energy distribution of accelerated electrons
(Paitanescu and Kumar 2001)
– p (power law index) ~ 2 to 3
• SED fits (GLAST+?)
E. do Couto e Silva
15/21
•
•
GLAST Science Lunch
Dec 1, 2005
Modeling Delayed Emission for GRB941017
High Energy data
constrains
– Total Energy
– Lorentz factor
– Ambient density
Two models used
– External shock
• e- from FS IC
scatters g from
RS
• SSA is
important
– Internal Shock
• Dt = 10-5
E = 3 × 1054 ergs, n = 0.10 cm-3, G i = 300, eB,r = 10-1, eB,f= 10-6, z = 0.10
E = 1 × 1055 ergs, n = 0.10 cm-3, G i = 200, eB,r = 10-3, eB,f= 10-5, z = 0.10
E = 1 × 1054 ergs, n = 0.03 cm-3, G i = 220, eB,r = 0.2, eB,f= 10-6, z = 0.06
E = 3 × 1052 ergs, n = 0.10 cm-3, G i = 1500, eB,r = 10-7, eB,f= 10-7, z = 0.15
Dt = 10-5
Data from Gonzalez et al 2003
GRB941017
LAT
Pe’er & Waxman 2004
E. do Couto e Silva
16/21
•
•
•
GLAST Science Lunch
Dec 1, 2005
Multiwavelength Observations to Constrain Models
Model
– Prompt emission from
internal shocks
• in relativistic wind
Pe’er & Wazman 2004
– Spectra as high as 10
GeV
• Flux has a strong
dependence on G
• Measure cut-offs with
GLAST!
Ratio of peaks in kev/GeV can
be used to constrain ratio of
eB / ee
• LAT + GBM?
• Swift + GLAST?
100 keV
eB = 0.33
1 GeV
l’ < 10
G = 600
Dt = 10-4
eB = 0.01
eB = 0.0001
Caveat:
– single shell collisions
E. do Couto e Silva
17/21
•
•
•
GLAST Science Lunch
Dec 1, 2005
Which Model do we Choose?
Region I
– Proton synchroton
• hard with GLAST
• large eB
Region II
– Inverse Compton
• good for GLAST
• denser medium
helps
Typical from afterglows
(Painatescu & Kumar 2001)
(Painatescu & Kumar 2002)
• small eB :internal
shocks
Region III
– Electron synchroton
• not in GLAST
range
E. do Couto e Silva
Zhang & Meszaros 2001
18/21
•
GLAST Science Lunch
Dec 1, 2005
We need DATA !
Region I
– Proton synchroton
• hard with GLAST
Zhang & Meszaros 2001
z = 0.1
z=1
940217
(Hurley 1994)
• large eB
•
Region II
– Inverse Compton
• good for GLAST
• denser medium
helps
•
Region III
– Electron synchroton
• not in GLAST
range
E. do Couto e Silva
eB = 10-4
ee = 0.5
19/21
•
•
GLAST Science Lunch
Dec 1, 2005
Can we constrain p and G?
Model
– Prompt emission from internal
shocks
• in relativistic wind
G = 600
G = 350
G = 200
300 keV
– To escape the system photons
must have
• opacity of HE g to pair
production with LE g < 1
• opacity to pair production < 1
– SSC dominates above 100 MeV
• Power law index > 2
• GLAST can constrain p
Guetta & Granot 2003
1 GeV
30 MeV
Dt = 10 ms
Dt = 1 ms
Dt = 0.1 ms
GRB940217:
– G= 600, Dt = 0.1 ms, Ep = 200 KeV
– Should we expect variability
smaller that 0.1 ms?
E. do Couto e Silva
20/21
GLAST Science Lunch
Dec 1, 2005
To B or Not to B….
• Magnetic Fields may play a key role in GRB formation
– Magnetic Poynting flux dominates instead of radiation
pressure
– Easier to transport energy
• No need of baryons (good for internal models)
– Narrower Peak Energy distribution
– Energy components
• Electromagnetic
Large value makes it hard to
EM
• Internal

develop strong external shocks
U

U
int
kin
• Kinetic (bulk)
(Paitanescu and Kumar 2001)
• This will be my next talk…
E. do Couto e Silva
21/21
GLAST Science Lunch
Dec 1, 2005
Back up
E. do Couto e Silva
22/21
GLAST Science Lunch
Dec 1, 2005
Efficiency Problem
E. do Couto e Silva
23/21
GLAST Science Lunch
Dec 1, 2005
GRB Peak Energy
E. do Couto e Silva
24/21
GLAST Science Lunch
E. do Couto e Silva
Dec 1, 2005
25/21
GLAST Science Lunch
Dec 1, 2005
GRB940217
E. do Couto e Silva
26/21
GLAST Science Lunch
E. do Couto e Silva
Dec 1, 2005
27/21
GLAST Science Lunch
Dec 1, 2005
Cooling Processes in the Fireball
• electron collides inelastically/elastically with atoms/ions of the gas
– Relevant only at lower energies
• electron emits bremsstrahlung g during collisions
• electron Compton scatters with g from radiation field
• electron is deflected by B field and emits g
E. do Couto e Silva
28/21