Transcript High Energy Emission in Extragalactic Nonblazar Sources

High Energy Emission in Extragalactic Nonblazar Sources
Chuck Dermer
U.S. Naval Research Laboratory
July 4, 2006
Multi-Messenger Approach to Unidentified Gamma-Ray Sources
Barcelona, Spain
Armen Atoyan
U. de Montréal
Markus Böttcher
Ohio University
Jim Chiang
UMBC/GSFC
Bob Berrington
University of Wyoming
Catalog of Established High Energy (> 100 MeV) Gamma-Ray Sources
Solar System:
1. Sun/Solar Flares (1)
Galaxy:
1. Pulsars (~8)
2. SNRs/Diffuse cosmic-ray induced
radiations (~10)
3. High-mass microquasars (2)
4. Pulsar Wind Nebulae and X-ray
Binaries (~dozen)
Extragalactic:
1. Diffuse CR emissions (LMC)
2. Blazars + Radio Galaxies (Cen A,
M87) (~100 + 2)
3. GRBs (~8)
3. Clusters of Galaxies?
4. Dark Matter Emission??
EGRET Unidentified Sources (~170)
HESS/TeV Unidentified Sources (>15)
GLAST Unidentified Sources (tbd)
Outline
Gamma Ray Bursts:
1.
Observations
Evidence for Multiple Components: Results from EGRET and BATSE
Rapid X-ray Declines Discovered with Swift
2.
3.
4.
5.
Blast Wave Model: Leptonic Processes
Blast Wave Model: Hadronic Processes
GRB/Cosmic Ray/g-ray/Neutrino Connection
SGRBs
Clusters of Galaxies:
1.
2.
3.
Merger and Accretion Shocks
Spectral Analysis
Predictions
1. Gamma Ray Bursts
subsecond variability
GRB 940217
Long (>90 min) g-ray emission
(Hurley et al. 1994)
GRB 940217
 Nonthermal
processes
Two components seen in
two epochs
MeV synchrotron and
GeV/TeV SSC
g-g lower limit to the
Two components seen in two separate epochs
bulk Lorentz factor
How to explain the two components?
G of the outflow
How to explain the two
components?
Anomalous High-Energy Emission Components in GRBs
Evidence for Second Component from BATSE/TASC Analysis
−18 s – 14 s
1 MeV
14 s – 47 s
47 s – 80 s
80 s – 113 s
Hard (-1 photon spectral
index) spectrum during
delayed phase
113 s – 211 s
GRB 941017
(González et al. 2003)
100
MeV
Second Gamma-ray Component in GRBs: Other Evidence
Atkins et al. 2002
Bromm & Schaefer 1999
(Requires low-redshift GRB to avoid attenuation
by diffuse IR background)
Delayed high-energy g-ray emission from superbowl burst
Seven GRBs detected with EGRET either during prompt MeV burst emission or
after MeV emission has decayed away (Dingus et al. 1998)
Average spectrum of 4 GRBs detected over 200 s time interval from start of
BATSE emission with photon index 1.95 (0.25) (> 30 MeV)
Swift Observations of Rapid
X-Ray Temporal Decays
Tagliaferri et al. (2005)
O’Brien et al. (2006)
Opacity Constraints: Lower Limits to G
GRBgg
940217
 Nonthermal
processes
Two components seen in
two epochs
MeV synchrotron and
GeV/TeV SSC
g-g lower limit to the
bulk Lorentz factor
G of the outflow
How to explain the two
components?
Nonthermal g-Ray Emission: gg Transparency Argument for
Bulk Relativistic Motion
In comoving frame, avoiding threshold condition for gg interactions
requires
 1  1; Peak Flux : 10-6 f -6 ergs cm -2 s -1
Requirement that gg optical depth be less than unity:
T 2 ' 2
ctv
 gg 
( ' )n ph ( ' ) rb , rb 

3 1
1
(1  z )
  200 [(1  z )d 28 ]
1/ 3
f -6 E (GeV ) 1 / 6
[
]
tv ( s )
Dermer, astro-ph/0402438
Baring 2006
Blast Wave Physics with Leptons
Electrons
•
•
•
Acceleration by Fermi Processes
Power in electrons and magnetic field determined by e and B parameters
Radiation and cooling by synchrotron and Compton Processes
External Medium
G
Unshocked shell
Cloud
Forward Shock
Reverse Shock
G0
Structured jet
*
GRB source
Colliding Shells
Captured
particle
Shocked shell
GeV/TeV Component from
Leptonic Processes
Observed properties sensitive to initial Lorentz
factor G0 of outflow (or baryon loading)
Dominant SSC component in some cases
Dermer, Chiang, and Böttcher (2000)
Blast Wave Physics with Leptons and Hadrons
Protons
• Acceleration by Fermi processes
• Energy content in protons determined by e, B parameters: p =1- e - B
• Radiative cooling by
1. Proton synchrotron
2. Photopair production
3. Photopion production
•
Escape from blast wave shell
p  B  p g
p  g  p  e  ep g  N 
Photopion Production
1.
2.
Resonance Production
D+(1232), N+(1440),…
Direct Production
pgn+
, pg
,
D++ -
Mücke et al. 1999
Threshold  m  150 MeV
pgD0+
3.
4.
Multi-pion production
QCD fragmentation models
Diffraction
Couples photons with r0, w
Er r
Two-Step Function Approximation for Photopion Cross Section
Atoyan and Dermer 2003
 ( Er )  340b, 200 MeV  Er  500 MeV , K1  0.2
120b, Er  500 MeV , K2  0.6
Kin ( Er )  ˆ  70b, Er  200 MeV
(useful for energyloss rate estimates)
Photopion Processes in a GRB Blast Wave
Threshold : g p   m  400
  h / me c
2
Threshold energy E p of protons interacting with photons
with energy pk (as measured by outside observer)
E p  mp c2Gg p
f  F
Fast cooling
f  pk
Describe F spectrum as a broken
power law
Protons with E > E p interact with
photons with  < pk, and vice versa
a= 1/2
4/3
3
abs c
b = (2-p)/2  -0.5
s=2
g0= gc
g1= gmin
 min
  pk
2

Photopion Energy Loss Rate in a GRB Blast Wave
Relate F spectrum to comoving photon density nph(´) for
blast-wave geometry (´2nph(´)dL2f/x2G2)
Calculate comoving rate t´-1f(Ep) = rf in comoving frame
using photopion (f) cross-section approximation
 Ep
rf
(0  a  1)
Kf
 Ep
All factors can be easily derived
from blast-wave physics (in the
external shock model)
1- a
1-b
Ep
Ep
Choose Blast-Wave Physics Model
Adiabatic blast wave with apparent total isotropic energy release
1054 E54 ergs (cf. Friedman and Bloom 2004)
Assume uniform surrounding medium with density 100 n2 cm-3
Relativistic adiabatic blast wave decelerates according to the
relation
(Böttcher and Dermer 2000)
Deceleration length
3 5 7
(Mészáros and Rees 1993)
Deceleration timescale
1 s 10 s
Why these parameters?
(see Dermer, Chiang, and Mitman 2000)
(Chiang and Dermer 1999)
Energies and Fluxes for Standard Parameters
Standard parameter set: z = 1
F flux ~ 10-6 ergs cm-2 s1
Duration ~ 30 s
Requires very energetic
protons (> 1015 eV) to
interact with peak of the
synchrotron spectrum
-1
Comoving Rates (s )
Characteristic hard-to-soft
evolution
10
Standard Parameters
1
r
-1
10
10
10
E (10 eV)
p
1/t'
10
18
acc
ava
f
-3
-6
E (MeV)
-5
r
pk
esc
r
-7
f
r
p,syn
-9
1
10
100
1000
Observer time t(s)
10
3
10
2
10
1
10
0
10
-1
10
-2
10
-3
10
-4
10
-5
10
-6
10
-7
Energies and fluxes
Epk ~ 200 keV at start of
GRB
10
Photopion Rate vs. Available Time for Standard Parameters
Standard parameter set: z = 1
-1
Comoving Rates (s )
Unless the rate is greater
than the inverse of the
available time, then no
significant losses
10
10
Standard Parameters
1
r
-1
1/t'
10
10
10
10
18
acc
E (10 eV)
p
ava
f
-3
-6
E (MeV)
-5
r
pk
esc
r
-7
f
r
p,syn
-9
1
10
100
1000
Observer time t(s)
10
3
10
2
10
1
10
0
10
-1
10
-2
10
-3
10
-4
10
-5
10
-6
10
-7
Energies and fluxes
Photopion rate increases
with time for protons
with energy Ep that have
photopion interactions
with photons with pk
Acceleration Rate vs. Available Time for Standard Parameters
Standard parameter set: z = 1
Implicitly assumes Type 2
Fermi acceleration, through
gyroresonant interactions in
blast wave shell
Makes very hard proton
spectrum n´(g´p)  1/g´p
Dermer and Humi 2001
-1
Comoving Rates (s )
Take zacc = 10: no problem
to accelerate protons to Ep
10
10
Standard Parameters
1
r
1/t'
10
10
10
10
18
acc
-1
E (10 eV)
p
ava
f
-3
-6
E (MeV)
-5
r
pk
esc
r
-7
f
r
p,syn
-9
1
10
100
1000
Observer time t(s)
10
3
10
2
10
1
10
0
10
-1
10
-2
10
-3
10
-4
10
-5
10
-6
10
-7
Energies and fluxes
Assume Fermi acceleration
mechanism; acceleration
timescale = factor zacc
greater than the Larmor
timescale t´L = mcg´p/eB
Escape Rate vs. Available Time for Standard Parameters
Standard parameter set: z = 1
Diffusive escape from blast
wave with comoving width
<x> = x/(12G).
No significant escape for
protons with energy Ep until
>>103 s
-1
Comoving Rates (s )
10
Standard Parameters
1
r
-1
1/t'
10
10
10
10
18
acc
E (10 eV)
p
ava
f
-3
-5
r
-6
E (MeV)
pk
esc
r
-7
f
r
p,syn
-9
1
10
100
1000
Observer time t(s)
10
3
10
2
10
1
10
0
10
-1
10
-2
10
-3
10
-4
10
-5
10
-6
10
-7
Energies and fluxes
Calculate escape timescale
using Bohm diffusion
approximation
10
Proton Synchrotron Loss Rate vs. Available Time
Standard parameter set: z = 1
Proton synchrotron energyloss rate:
-1
No significant proton
sychrotron energy loss for
protons with energy Ep
10
Standard Parameters
1
r
-1
1/t'
10
10
10
18
acc
E (10 eV)
p
ava
f
-3
-6
E (MeV)
-5
r
pk
esc
r
-7
f
r
p,syn
10
-9
1
10
100
1000
Observer time t(s)
10
3
10
2
10
1
10
0
10
-1
10
-2
10
-3
10
-4
10
-5
10
-6
10
-7
Energies and fluxes
Comoving Rates (s )
10
Gamma-Ray Bursts as Sources of High-Energy Cosmic Rays
Solution to Problem of the Origin of Ultra-High Energy Cosmic Rays
(Waxman 1995, Vietri 1995, Dermer 2002)
Hypothesis requires that
GRBs can accelerate cosmic
rays to energies > 1020 eV
Injection rate density
determined by GRB
formation rate (= SFR?)
GZK cutoff from photopion
processes with CMBR
Pair production effects for
ankle
(Berezinsky and Grigoreva 1988,
Berezinsky, Gazizov, and Grigoreva 2005)
(Wick, Dermer, and Atoyan 2004)
Rates for 1020 eV Protons
Standard parameter set: z = 1
-2
10
Calculated at E =10
p
1/t'
-3
-1
Comovin Rates (s )
For these parameters, it
takes too long to accelerate
particles before undergoing
photopion losses or
escaping.
10
r
10
eV
ava
r
-4
20
esc
acc
-5
r
10
r
-6
10
f
p,syn
-7
10
1
10
100
1000
Observer time t(s)
10
4
Rates for 1020 eV Protons with Equipartition Parameters
Equipartition parameter set with density = 1000 cm-3, z = 1
10
-2
r
Calculated at E =10
20
p
esc
eV
-1
Comovin Rates (s )
Within the available time,
photopion losses and
escape cause a discharge of
the proton energy several
hundred seconds after GRB
10
-3
r
r
10
-4
10
-5
1
1/t'
acc
ava
p,syn
r
10
100
Observer time t(s)
1000
f
Rates for 1020 eV Protons with Different Parameter Set
New parameter set with density = 1000 cm-3, z = 1
Escape from the blast wave
also allows internal energy
to be rapidly lost (if more
diffusive, more escape)
10
-2
Calculated at E =10
20
1/t'
-1
Comoving Rates (s )
p
10
ava
-3
r
acc
r
10
-4
r
r
10
f
esc
p,syn
-5
1
10
100
Observer time t(s)
1000
eV
Blast Wave Evolution with Loss of Hadronic Internal Energy
Assume blast wave loses 0, 25, 50, 75, 90, and 95% of its energy at x =
6x1016 cm.
Transition to radiative solution
Rapid reduction in blast wave
Lorentz factor G = (P2 +1)1/2
Rapid decay in emissions
from blast wave, limited
by curvature relation
Highly radiative phase---due to
escape of UHECRs from GRB
blast wave---proposed as
explanation of Swift observations
of rapid X-ray declines in
GRB light curves
Photon and
Neutrino Fluence
during Prompt
Phase
Nonthermal Baryon
Loading Factor fb = 1
Ftot = 310-4 ergs cm-2
 = 100
Requires large baryon load
to explain GRB 941017
Hard g-ray emission component from hadronic cascade radiation
inside GRB blast wave
Second component from outflowing high-energy neutral beam of
neutrons, g-rays, and neutrinos


pg    e ( n, p, )
  0  2g  e
gg Optical Depth
Photon attenuation strongly dependent on  and tvar in collapsar model
F tot  3  10-4
ergs cm -2 ,
50 one sec
pulses
gg evolves
in collapsar
model due to
evolving
Doppler factor
and internal
radiation field
Neutrinos from GRBs in the Collapsar Model
requires Large Baryon-Loading
Nonthermal Baryon Loading Factor fb = 20
(~2/yr)
Dermer & Atoyan 2003
Rapidly Declining X-ray Emission Observed with Swift
F   -
Zhang et al. 2006
Rising phase of light curve shorter than declining phase in
colliding shell emission
Difficult for superposition of colliding-shell emissions to explain
Swift observations of rapid X-ray decay
Rapid X-ray Decays in Short Hard Gamma-Ray Bursts
GRB 050724
Barthelmy et al. (2005)
Loss of internal energy through ultra-high energy particle escape:
UHECRs from SGRBs?
 High-energy g-rays expected from SGRBs from leptonic and, possibly,
hadronic components
Implications and Predictions
• Photopion production
Decay lifetime  900 gn seconds
Cascade radiation, including proton synchrotron radiation, forms a new g-ray
emission component: Explanation of GRB 940217, GRB 941017,…
Escaping neutrons and g-rays form hyper-relativistic electrons; transient gray/X-ray synchrotron halos, as in blazars (Coppi, Aharonian & Völk 1994)
• Unidentified g-ray Flashes: Proton synchrotron radiation
– Discover with GLAST or Milagro
– Need rapid alert from GLAST to TeV telescopes
2. Nonthermal Particles and Radiation Produced by
Cluster Merger Shocks
Thermal bremsstrahlung X-ray Emission of galaxy clusters traces gravitational well
Rich clusters (thousands of Galaxies;
~1015 Msun; kT ~ 5-10 keV, LX ~ 1043 1045 ergs s-1)
Velocity dispersions ~500-1000 km s-1
Poor clusters (hundreds of Galaxies;
~1014 Msun; kT ~ 1-5 keV, LX ~ 1041 1043 ergs s-1 )
Velocity dispersions ~250-500 km s-1
~5-10% of total mass of cluster; Orbital motion
dominated by distribution of dark matter
Which clusters are GLAST/TeV-bright?
Structure Formation
• Density fluctuations cause region to
collapse.
– Magnitude of the density fluctuation
determines the formation time
– Larger structures form by accreting
smaller clumps--hierarchical merging
– Lumpy, continuous accretion
Cluster Merger
• Simulation of merging clusters of galaxies
Shocks in Merging Clusters
• (0, R, ) (mass, curvature, and dark
energy)= (0.3, 0.0, 0.7)
– Redshift of cluster:
– Cosmic Microwave Background
(CMBR) dependence
• UCMBR(z) = UCMBR(z=0) (1 + z)4
• Rich clusters form by
accreting poor clusters
• Shocks in Merging Clusters
Particle Injection
• Power law distribution with exponential
cutoff Q ( E , t )  Q  ( pc)  exp - E 
-
0
e, p
e, p







Emax (t ) 
– Occurs only if M  1.0
– Occurs only during lifetime of shock
• Normalization

Emax
Emin
1

e
Ee, pQe, p E , t  dE  e, p  nIC M He
m p vs2   As vs 
2

– Where e,p is an efficiency factor, and is set to 5%.
– Typical values are Etot1063-64 ergs
Particle and Photon Energy Spectra: Coma Cluster
Fit to Data for the Coma Cluster
Galaxy Cluster Nonthermal Brightness
Nonthermal Emission from Cluster Merger Shocks
• Unidentified EGRET sources: Doubtful
• Diffuse Extragalactic g-ray Background: Few % contribution
Summary
Clusters of Galaxies
Unidentified EGRET sources: Doubtful
Diffuse extragalactic g-ray background: Few % contribution
Predictions: Handful (~ 5 – 10) detected with GLAST
(Merger vs. accretion shocks)
(Merger shock acceleration vs. turbulent acceleration)
GRBs
Highly radiative phase from UHECR escape in blastwave evolution
proposed to explain rapid X-ray declines in Swift GRB light curves
Predictions:
1.
2.
3.
Hadronic g-ray light consisting of cascading photopion and proton
synchrotron radiation varying independently of leptonic synchrotron
Strong GeV-TeV radiation and/or ultra-high energy (>1017 eV) neutrinos
correlated with rapidly decaying X-ray emission
UHECR emissivity following the GRB formation rate history of the universe
Back-up Slides
Synchrotron and SSC Radiation
Strong dependence of GRB emissions on G0
Selection bias to detect GRBs with Epk within
waveband of detector
Dominant SSC component in some cases
Chiang and Dermer (1999)
Two-Step Collapse (Short-Delay Supranova)
Model
1.
2.
3.
4.
5.
6.
Standard SNIb/c (56Ni production)
Magnetar Wind Evacuates Poles
GRB in collapse of NS to BH
Prompt Phase due to External Shocks with
Shell/Circumburst Material
Standard Energy Reservoir (NS collapse to BH)
Delay time ~<
Beaming from mechanical/B-field collimation
1 day (GRB 030329)
Infall Velocity