Transcript GRB prompt emission

GRB
Theory and observations
Useful reviews:
Waxman astro-ph/0103186
Ghisellini astro-ph/0111584
Piran astro-ph/0405503
Meszaros astro-ph/0605208
Gehrels 2009 ariv:0909.1531
Useful links:
http://qso.lanl.gov/~clf/papers (Chris Fryer lectures)
GRBs most luminous
objects in the Universe!!
• Sun Luminosity L~4 1033 erg/s
• Supernova L~1051 erg/s
• Galaxies with nuclei L~1048 erg/s
• GRB luminosity L~1052 erg/s
GRB light curves
GRBs: flashes of 0.1 MeV gamma rays that last 1-100 s
-ray observations summary
• Isotropy in the sky
• Duration: T90
• Variability:
0.2 s short
20 s long
Most show
Some
t ~ 64 ms
t ~ 1 ms
• Flux: f = 10-4 -10-7 erg/cm2 s
• Rate R
300/yr BATSE and 100/yr Swift
GRB
3 July 1969: first detection of a GRB by
Vela 5A
Vela Satellites
• 105 km Orbits
• Launched in
pairs –
launched
1963-1965
• Operated
until 1979
• All satellites
allowed for
some
localization.
First Detected Gamma-Ray Burst
Vela Satellites - Results
• 73 Bursts in Gamma-Rays
over 10 years
• Not from the Earth (not
weapons tests) and not in
the plane of solar system
Ray Klebasadel
Gamma-Ray Bursts in the
Solar System
• Lightning in the
Earth’s
atmosphere
(High Altitude)
• Relativistic Iron
Dust Grains
• Magnetic
Reconnection in
the Heliopause
Red Sprite Lightning
Gamma-Ray Bursts in the Milky
Way
• Accretion Onto White
Dwarfs
• Accretion onto neutron
stars
I) From binary companion
X-ray Novae
II) Comets
• Neutron Star Quakes
• Magnetic Reconnection
Galactic Gamma-Ray Bursts:
Soft Gamma-Ray Repeaters
One Class of GRBs
Is definitely Galactic:
Soft gamma-ray
Repeaters (SGRs)
Characteristics:
1) Repeat Flashes
2) Photon Energy
Distribution lower
Energy than other
GRBs (hard x-rays)
X-ray map of N49 SN remnant. The white
Box shows location of the March 5th event
Models for SGRs
• Accretion
I) Binary
Companion - no
companion seen
II) SN Fallback –
Too long after
explosion
• Magnetic Fields
~1015 G Fields “Magnetars”
Extragalactic Models
• Large distances
means large energy
requirement
(1051erg)
• Event rate rare (106-10-5 per year in an
L* galaxy) – Object
can be exotic
Cosmological Models
• Collapsing WDs
• Stars Accreting on
AGN
• White Holes
• Cosmic Strings
• Black Hole Accretion
Disks
I) Binary Mergers
II) Collapsing Stars
Black-Hole Accretion
Disk (BHAD) Models
Binary merger or
Collapse of rotating
Star produces
Rapidly accreting
Disk (>0.1 solar
Mass per second!)
Around
black
hole.
Massive
Star
Collapse
Collapsar Model – Collapse of a Rotating
Massive Star into a Black Hole
Main Predictions: Beamed Explosion,
Accompanying supernova-like explosion
Stan Woosley
BATSE - Burst And Transient Spectrometer Experiment
BATSE Module
8 Detectors
Almost Full Sky Coverage
Few Degree Resolution
20-600keV
BATSE Consists of
two NaI(TI) Scintillation
Detectors: Large
Area Detector (LAD)
For sensitivity and the
Spectroscopy Detector
(SD) for energy coverage
Galactic Models
BATSE Results – Isotropy
Cosmological Models Favored!
Gamma-Ray Burst Lightcurves
GRB990316
GRB Lightcurves have
A broad range of
Characteristics
Fast Rise Exponential Decay
“FREDs”
GRB970508
Gamma-Ray Burst Durations
Two Populations:
Short – 0.03-3s
Long – 3-1000s
Possible third
Population
1-10s
Gamma-Ray Burst Duration
vs. Energy Spectrum
BATSE - Summary
• GRBs are Isotropic – The beginning of
the end for Galactic Models, but
persistent theorists move the Galactic
Models to the Halo
• GRBs come in all shapes and sizes but
two obvious subgroups exist I) Short, Hard Bursts
II) Long, Soft Bursts
BeppoSAX
Italian-Dutch Satellite
Launch: April 30, 1996
Goal: Positional Accuracy
<5 arc minutes
Honoring Giuseppe Occhialini
High Pressure Gas Scintillation
Proportional Counter
WFC – 40o x 40o, 2-28keV
BeppoSAX Instruments
LECS/MECS
• Xenon Gas
Scintillator
• Energy Range: .11keV (1-10keV)
• ~1 arc minute
resolution
• Goal – Localize
Object
HPGSPC PDS
• HPGSPC - High Pressure
Xenon/He Gas
• PDS Phoswitch - NaI(Tl),
CsI(Na) Scintillators
• 4-120keV (15-300keV)
• Goal – Broad Energy
resolution in X-ray narrow
field
BeppoSAX: I GRB sono sorgenti a
distanze cosmologiche!
Costa+ 1997 BeppoSAX
Pedichini+ 1997 Campo imperatore
Van Paradijs+ 1997 WHT
GRB 970228 – host galaxy
observed?
This blob, a peculiar
Galaxy to be sure,
Is in the same position
As the Burst!
Could it have been the
GRBs host?
The galaxy has a
Redshift of 0.695.
GRB 970508 – Optical
Counterpart
BeppoSAX
X-ray
Localization
Allowed a
The Optical
Transient to
Be detected
While still on
The rise.
OT allowed
Spectral
Measurement!
Metzger et al. 1997
flux
GRB970508 – Absorption Lines:
z=0.835
Wavelength
Optical Emission
Absorption
Fe II
Fe II
Mg II I
flux
Mg II
Wavelength
Host Galaxy Detected
for GRB970508
flux
Z=0.835
Wavelength
Radio Twinkling can also be used to estimate
the GRB distance: consistent with z=0.835
Just as the Earth’s
Atmosphere
Causes light
To scatter
Causing point
Sources to
“twinkle”, the
Interstellar
Medium causes
Radio emission
To twinkle. When
The burst gets
Large enough,
Like planets, the
Twinkling stops.
Waxman, Kulkarni, & Frail 1997
T=0, point
Source
T=t, r=c t
Where c is speed of light
ISM Scattering
Twinkle,
Twinkle
Observer
Always Sees
Part of Burst
A crash Course in
Scintillations
Scintillations determine the size of the source in a
model independent way. The size (~1017cm) is in a
perfect agreement with the prediction of the
Fireball model.
GRB971214 @z=3.42
GRB NH and AV
HETE2
Fregate: 6-400 keV GRB triggers and
low res. Spectra
WXM 2-25 keV, medium energy
resolution and 10arcmin localization
SXC 0.5-10 keV, good energy
resolution and 1arcmin localization
Swift: a new era for GRB studied
Burst Alert Telescope (BAT)
- 32,000 CdZnTe detectors
- 2 sr field of view
X-Ray Telescope (XRT)
- CCD spectroscopy
- Arcsec GRB positions
UV-Optical Telescope (UVOT)
- Sub-arcsec position
- 22 mag sensitivity
Spacecraft slews XRT &
UVOT to GRB in <100 s
Swift GRBs
XRF
Short
GRB
XRF
XRF
XRF
XRF
Short
GRB
XRF
Short
GRB
Short
GRB
Short
GRB
XRF
XRF
Short
GRB
XRF
Short
GRB
XRF
XRF
Short
GRB
Short
GRB
Swift localizes short GRBs
XRT
BAT
• elliptical hosts
• low SF rates
• offset positions
• redshifts z ~ 0.2
>> inconsistent with
collapsar model
>> supportive of
NS-NS model
XRT
Chandra
Il GRB piu’ lontano, quello piu’
brillante e quello piu’ energetico
GRB080913
GRB080916C Fermi -rays
GRB080319B
3 GRB @ z>6
GRB050904
Ly break in the IR
J=17.6 at 3.5 hours
Subaru Spectroscopy
Observational Constraints on the
Central Engine
•
•
•
•
•
•
Host Galaxies
GRB Environments
Prompt Emission
Bumps in the Afterglow (SN?)
Energetics and Beaming
Using GRBs as Cosmological Probes
I: Host Galaxies
Accurate positions
Allowed Astronomers
To watch the bursts
Fade, and then
Study their Host
Galaxy!
The fading optical
afterglow of GRB 990123
as seen by HST
on Days 16, 59 and 380
after the burst.
Host Galaxy
Optical Afterglow
Properties
Of Host
Galaxies
I) Like Many
Star-forming
Galaxies
At that
Observed
redshift
Holland 2001
II) Star-formation rates high, but consistent
With star forming galaxies.
Location, Location, Location
(In addition to detecting hosts, we can determine where
a burst occurs with respect to the host.
GRB hosts
• GRBs trace brightest regions
in hosts
• Hosts are sub-luminous
irregular galaxies
 Concentrated in regions of
most massive stars
 Restricted to low metallicity
galaxies
Distribution
Follows
Stellar
Distribution
If we take
These
Positions
At face
Value,
We can
Determine
The
Distribution
Of bursts
With respect
To the halfLight radius
Of host
Galaxies!
This Will
Constrain
The models!
GRB Hosts Exhibit Larger Mg
line Equivalent Widths Than
QSO absorbers: Higher Density?
Fiore 2000
Salamanca et al.
2002
Savaglio, Fall &
Fiore 2003
Results from low resolution spectroscopy
High dust
depletion
High dust content
Denser clouds
Savaglio, Fall & Fiore 2003
2) Metallicity depends on galaxy mass
Savaglio et al. 2008
Berger et al. 2006
Star-formation rate in GRB hosts
Savaglio+ 2008
What we’ve learned from GRB
Hosts!
• Hosts of long GRBs are star-forming
galaxies
• GRBs trace the stellar distribution (in
distance from galaxy center)
• GRBs occur in dense environments
(star forming regions?)
Using GRBs as Cosmological
Probes
Gamma-Ray Bursts are observed at extremely
high redshifts and can be used to study the early
universe.
• Star Formation History
• Beacons to direct large telescopes to study
nascent galaxies
• Studies of intervening material between us and
GRB – akin to quasar absorption studies
METAL ABUNDANCES IN HIGH z GALAXIES
GRB explosion
site
Circumburst
environment
Host gas
far away
To
Earth
Redshift
Distribution
Of GRBs
With known
Redshifts
(2002)
Redshifts
As high as
5 observed!
Lloyd-Ronning et al. 2002
Lloyd-Ronning, Fryer, & Ramirez-Ruiz 2002
Solid squares
Denote bursts
With observed
Redshifts.
Open squares
Denote
Positions using
A LuminosityVariability
Relation.
(Fenimore &
Ramirez-Ruiz
2000).
Dashed line
Artifact of
Luminosity
Cut-off in FR-R
Sample.
Redshift distributions
Pre-Swift
Swift
0
1 2 4
Redshift (z)
13
Galaxies
Quasars
GRBs
10
12
10
8
High resolution spectroscopy: GRB021004
FORS1 R~1000
CIV
CIV
z=2.296
z=2.328
UVES R=40000
z=2.296
z=2.328
GRB050730 UVES spectrum
GRB locations within galaxies
GRBs show higher gas densities and metallicities,
And have significantly lower [(Si,Fe,Cr)/Zn] ratios,
Implying a higher dust content: Star Formation Region
History of metal enrichment
000926
050820
050401
060206
030323
050904
050730
Savaglio+2003
Prochaska+ 2003
GRB host galaxy metallicities
However… metallicity depends on:
1)Impact factor
2)Galaxy mass
3)Star-formation rate
4)Etc….
1) Metallicity depend on impact factor
GRB021004
Variability
GRB060418 z=1.49
VLT/UVES
Vreeswijk et al. 2007
Intervening absorbers
Ly forest: deviation from what is already known from
quasar forests. ``Proximity effect'' should be much reduced for
GRBs. An accurate determination of dn/dz at high z has strong
implications for investigations of the re-ionization epoch, since
the optical depth due to Ly line blanketing is evaluated by
extrapolating the Ly dn/dz measured at lower-z.
MgII and CIV absorbers: Incidence of MgII absorbers
~4 times higher than along QSO sight-lines. Incidence of CIV
absorbers similar… WHY???
Dust composition/evolution
the case of GRB 050904 @z=6.3
Large X-ray absorption
and UV dust extinction
Haislip WFCAMUKIRT ~0.5 days,
Ly corr. = 3.02
Tagliaferri FORSVLT ~1 day, Ly
corr. = 1.27
Haislip GMOSGemini ~3 days,
Ly corr. = 2.38
GRB 050904 z=6.3
Stratta et al 2007
[email protected] extinction
curve
0.5 day A3000=0.89+\-0.16
1 day A3000=1.33+\-0.29
3 days A3000=0.46+\-0.28
NH~1023 cm-2
 AV/NH~50 times lower
than Galactic!!
@z~6 no dust from AGB
stars. Only sources are
CCSNe (and AGNs)
Much less dust and much smaller AV/NH
GRB Environments II:
Studying the environment using radio
and optical observation of GRBs
• Density profiles are different for different
environments: massive stars will be
enveloped by a wind profile.
• These different density profiles produce
different radio, optical emission.
The Density Profile from Winds
ISM density is constant
The
Shock
Radius
Depends
On the
Density
Profile!
Radio
And
Optical
Light
Curves
Are a
Function
Of this
Radius!
GRB021004
For Many
Gamma-Ray
Bursts,
Wind-swept
Environments
Best fit the
Data (radio
And R-band
Data best
Diagnostics!
Roger Chevalier
Li & Chevalier 2003
On the Surface,
It appears we
Can constrain
The environments,
But, beware,
There still remain
Many free
Parameters in
These calculations!
The connection between SNe and
GRBs
Afterglow and GRB Energetics
IV:
• As we learned yesterday, afterglows allowed
us to calculate redshifts.
• Assuming a cosmology, we can then get
distances.
• Assuming isotropic explosions, we can
estimate the GRB energies! These energies
range over many orders of magnitude.
GRB Redshifts (2000)
GRB
Redshift
Isotropic
Energy
GRB
Redshift
Isotropic
Energy
GRB970228 0.695
5x1051
GRB990308 >1.2?
NA
GRB970508 0.835
8x1051
GRB990506 1.3
NA
GRB970828 0.958
NA
GRB990510 1.619
3x1053
GRB971214 3.418
3x1053
GRB990705 0.86
NA
GRB980326 1?
3x1051
GRB990712 0.430
NA
GRB980329 2 or 3-5
NA
GRB991208 0.706
1.3x1053
GRB980425 0.0085
1048
GRB991216 1.02
6.7x1053
GRB980613 1.096
NA
GRB000131 4.5
1054
GRB980703 0.966
1x1053
GRB000418 1.118
5x1052
GRB990123 1.600
3x1054
GRB000926 2.066
2.6x1053
Afterglow and GRB Energetics
• As we learned yesterday, afterglows allowed
us to calculate redshifts.
• Assuming a cosmology, we can then get
distances.
• Assuming isotropic explosions, we can
estimate the GRB energies! These energies
range over many orders of magnitude.
• But are GRBs isotropic?
Jet Signatures
Eg = (1 - cos q b )Eg ,iso
f b = (1 - cos q b ) » q b /2
2
é n ohg ù
qb µ ê
ú
E
ëê g ,iso ûú
1/ 8
é t break ù
ê 1+ z ú
ë
û
3/8
Stanek et al. (2001)
GRB 010222
Energy and Beaming
Corrections
•
15 events with z and t_jet
•
•
The dispersion in isotropic
GRG energies results from a
variation in the opening (or
viewing) angle
The mean opening angle is
about 4 degrees (i.e. fb-1 ~
500 )
Geometry-corrected energies
are narrowly clustered
(1=2x)
Eg = 5 ´1050 erg
(for n o = 0.1 cm -3 assumed)
Frail et al. (2001)
Energy and Beaming
(Continued)
24 events with
z and t_jet
•
•
Outliers
Improved analysis
• Larger sample
• Used measured densities
• Error propagation
Geometrically corrected
gamma-ray energy …
Eg = 1.3 ´1051 erg
•
•
Bloom, Frail & Kulkarni (2003)
Increase is due to using
real density values
1 of 0.35 dex (2.2x)
Summary of GRB Energetics
• Gamma-ray bursts and
their afterglows have
(roughly) standard
energies
Eg » 1051 erg
E k » 1051 erg
• Robust result using
several complementary
methods
E  gamma-rays
Ek  X-rays
Ek  BB modeling
Ek  Calorimetry
E shock » Eg + E k » few ´1051 erg
hg º Eg / E shock » 0.5
SN/GRB connection!
GRBs have SN-like outbursts.
But these bursts are beamed, and we won’t see all
explosions as a GRB.
What do we make of the SN/GRB connection:
I)
All GRBs produce SNe?
II) All SNe are GRBs (only those observed along the jet
axis are GRBs)?
Are either of these true?
Ambitious Theorists – New SN
Mechanism
• Collapsar Theorists
argue I) is true, but not
II)
• Others argue that all
supernovae have jets
(e.g. asymmetries in
SN1987A) and the
standard SN engine is
wrong!
• SN-like is NOT SN
What fraction of SNe are GRBs?
The GRB community tends to not talk to the
SN community. Hence this problem has
lingered for a long time. The simple fact is
that the SN-like spectra and lightcurves are
quite different than true SNe.
But let’s assume we don’t know this, how else
can we tell? - Radio!
A Complete Radio Catalog
• 5 yr period (1997-2001)
• BeppoSAX, IPN, RXTE and
HETE satellites
• 75 GRBs searched for radio
AGs
• searches at 5 and 8.5 GHz
• frequencies 0.8-650 GHz
• 1521 flux density
measurements (or limits)
• 2002-2003 data on Web
Frail, Kulkarni, Berger and Wieringa AJ May 2003
http://www.aoc.nrao.edu/~dfrail/grb_public.shtml
Cumulative Flux Density
Distribution
50 %
• Max radio flux 2 mJy
• 19 detections
– mean=315+/-82 uJy
• 44 GRB in total
– mean = 186+/-40 uJy
• 50% of all bursts are brighter
than 110 uJy
• Radio afterglow observations
are severely sensitivity
limited!
Complete sample of 44 GRBs with 8.5 GHz
measurements made between 5 and 10 days post-burst
Spectral Radio Luminosity
Ln = 4pd 2L Fn (1 + z )a - b -1
where Fn µ t an b
mean » 1031 erg s -1 Hz -1
SN1993J = 2 ´ 10 27 erg s -1 Hz -1
n g Fg
» 2 ´108 !
n R FR
Complete sample of 18 GRBs with redshifts and 8.5 GHz
measurements made between 5 and 10 days post-burst
Fireball Calorimetry
Frail, Waxman & Kulkarni (2000)
E o » 5 ´1050 erg
n o » 1 cm -3
• Long-lived radio
afterglow makes a
transition to NR
expansion
– no geometric uncertainties
– can employ robust Sedov
formulation for dynamics
– compare with equipartition
Most energy estimates
require knowledge of the
geometry of the outflow
– radius and cross check
with ISS-derived radius
• Limited by small
numbers
How Common are EnginePowered SNe?
VLA/ATCA survey of 34 Type
Ib/c SNe to detect off-axis
GRBs via radio emission
Berger PhD
• Most nearby SNe Ib/c do not have
relativistic ejecta
• Two distinct populations
• Ek(GRB)<<1 foe (hydo
collapse)
• <10% are 1998bw-like