Download GRB EXPERIMENT

SOFT GAMMA REPEATERS
AN OBSERVATIONAL REVIEW
SOFT GAMMA REPEATERS
Kevin Hurley
UC Berkeley
Space Sciences Laboratory
Kevin Hurley
UC Berkeley
Space Sciences Laboratory
[email protected]
THE SOFT GAMMA REPEATERS ARE SPORADIC
SOURCES OF BURSTS
• SGRs can remain dormant for many years; during these periods,
no bursting behavior is observed
• They become active and emit bursts at apparently random times
• Two common types of bursts:
– Short (100 ms, up to 1041 erg s-1)
– Giant flares (several hundred seconds, periodic emission, up to
1046 erg s-1)
BURSTING ACTIVITY OF 3 SGRs OVER 17 YEARS
80
60
SGR1627-41
40
BURSTS PER 10 DAY INTERVAL
20
60
0
40
SGR1900+14
20
80
0
60
SGR1806-20
40
20
0
1992
1996
2000
YEAR
2004
2008
SINGLE, ~100 ms LONG BURST (MOST COMMON)
9
0
8
0
7
0
6
0
U
L
Y
S
S
E
S
S
G
R
1
9
0
0
+
1
4
2
5
1
5
0
k
e
V
9
8
0
5
3
0
A
5
0
COUNTS/32ms.
4
0
3
0
2
0
1
0
0
01
02
03
04
05
06
0
T
I
M
E
,
S
E
C
O
N
D
S
TWO SGR GIANT FLARES
COUNTS/0.5 s
10 4
5.16 s period
10000
Counts/0.5 s
SGR1900+14
SGR1900+14
AUGUST
1998
AUGUST
27271998
ULYSSES
ULYSSES
25-150 keV
25-150 keV
Eγ=4x1044 erg
10 5
SGR1806-20
DECEMBER
27, 2005
SGR1806-20
DECEMBER
27, 2004
RHESSI
RHESSI
20-100
20-100
keV keV
Eγ=8x1045 erg
7.56 s period
1000
100
10 3
10
50.0
100.0
150.0
200.0
250.0
TIME, s.
300.0
350.0
0
100
200
Time, s
Three phases:
1) Fast rise (<1 ms)
2) Very intense initial spike, ~100 ms long
3) Periodic decay (~300 s)
300
400
GIANT SGR FLARES ARE SPECTACULAR!
• Occur perhaps every 30 years on a given SGR
• Second only to supernovae in intensity
• Intense (Eγ≳1046 erg at the source, 1 erg/cm2 at Earth)
• Very hard energy spectra (up to >10 MeV)
• Create transient radio nebulae
• Cause dramatic ionospheric disturbances
• Should be detectable in nearby galaxies
SHORT BURST ENERGY SPECTRA, 2-150 keV:
SUM OF TWO BLACKBODIES, kT=3.4 and 9.3 keV
(SGR1900+14, Feroci et al. 2004)
BeppoSAX MECS
kT=3.4 keV
R= 14 km at 10 kpc
BeppoSAX PDS
kT=9.3 keV
R=2 km at 10 kpc
GIANT FLARE ENERGY SPECTRUM:
175 keV BLACKBODY, THEN 10 keV BLACKBODY
100
kT, keV
SPECTRAL TEMPERATURE
10
1
SGR1806-20 GIANT FLARE
RHESSI
Counts/0.5 s
10000
1000
100
10
0
100
200
Time, s
300
400
2
FLUX, photons/cm s keV
10
5
10
4
10
3
10
2
10
1
GIANT FLARE
SMALL BURST
10
10
0
-1
10
1
2
10
ENERGY, keV
3
10
THE SGRs ARE QUIESCENT X AND γ-RAY SOURCES
• Luminosities: 1034 – 1036 erg/s (>spin-down energy)
• This quiescent component varies slowly, and exhibits
pulsations (~10-20% pulsed fraction)
QUIESCENT X-RAY SOURCE ASSOCIATED WITH
SGR1806-20
ASCA, 2-10 keV
10-11 erg cm-2 s-1
INTEGRAL-IBIS, 18-60 keV
10-10 erg cm-2 s-1
QUIESCENT X-RAY FLUX LEVEL IS RELATED TO
THE BURSTING ACTIVITY
SGR1806-20
Woods et al. 2006
GIANT FLARE
Bursts and quiescent emission are probably both related
to magnetic stresses on the surface of the neutron star
P and P-dot FROM QUIESCENT SOFT X-RAYS (2-10 keV)
.
SGR1900 (P=5.16 s, P=10
Hurley et al. 1999
10-10 s/s
Woods et al. 1999
-10
s/s)
.
SGR1806 (P=7.48 s, P~10
-10
s/s)
Kouveliotou et al. 1998
8x10-11 s/s
Woods et al. 2000
SPINDOWN IS IRREGULAR, SOMETIMES RELATED TO
BURSTING ACTIVITY, SOMETIMES NOT RELATED
(Woods et al. 2002, 2006)
SGR1900+14
Woods et al. 2006
GIANT FLARE
This argues against accretion as the cause of the bursts
BROADBAND QUIESCENT X-RAY SPECTRA
Blackbody < 10 keV, Power Law > 20 keV
SGRs
AXPs
Götz et al. 2006
.
MAGNETARS COMPARED TO OTHER NS: P-P DIAGRAM
SGRs,
AXPs
High B
Radio
Pulsars
Radio
Pulsars
Millisecond
Radio
Pulsars
V. Kaspi 2006
ESSENTIAL SGR PROPERTIES
Giant
Flare?
P
s
.
P
s/s
1-10 keV
luminosity
erg/s
SGR1806-20
Dec 27
2004
7.46
~10-10
2x1035
SGR1900+14
Aug 27
1998
5.16
~10-10
3x1034
B,
Gauss
1/2
 3c 3 IP . 
P

 8 2R6 


8x1014
2-8x1014
SGR0525-66
Mar 5
1979
8
~7x10-11
1036
7x1014
SGR1627-41
No
2.6
1.2x10-11
1035
2x1014
SGR0501+45
No
5.8
5x10-12
1034
1014
1E1547-5408*
No
2.1
2.3x10-11
1033
2.2x1014
*initially thought to be an AXP
HOSTS AND PROGENITORS
• One or two SGRs are probably in supernova remnants
• One SGR may have been ejected from its supernova remnant
• Two SGRs are probably in massive star clusters
• The SNR association implies a normal progenitor mass (~5-8 M)
• The massive cluster association implies a massive progenitor (~50M)
THE SGR-SNR CONNECTION
•
SGR0525-66 is almost certainly
in the N49 SNR in the LMC
•
1E1547-5408 lies within the radio
SNR G327.24-0.13
•
SGR0501-4516 may have been
ejected from its supernova
remnant
MASSIVE CLUSTER-SGR ASSOCIATIONS
SGR1900+14
• ~13 stars
• 1-10 Myr old
(Vrba et al. 2000)
SGR1806-20
• ~12 stars
• 3-5 Myr old
• SGR progenitor mass ~48M
(Fuchs et al. 1999; Bibby et al. 2008)
COUNTERPARTS
• Two SGRs exhibited transient radio nebulae after giant flares
• One SGR has a persistent radio counterpart
• One SGR has a variable NIR counterpart
SGRs ARE TRANSIENT RADIO SOURCES AFTER GIANT
FLARES: RADIO NEBULA CREATED BY GIANT FLARE FROM
SGR1806-20 (Taylor et al. 2005)
VLA
THIS RADIO EMISSION COMES FROM AN EXPANDING CLOUD
OF RELATIVISTIC ELECTRONS ACCELERATED IN THE MAGNETOSPHERE
AND EXPELLED. BUT SGRs ARE NOT OBSERVABLE QUIESCENT RADIO SOURCES
SGR1806-20 IS INVISIBLE IN THE OPTICAL (nH~6x1022 cm-2),
BUT IT IS JUST BARELY VISIBLE IN THE INFRARED
10″
1.5″
mKs=20
m K’ =22
Kosugi et al. 2005
Israel et al. 2005
This is the only optical or IR counterpart to an SGR so far
IR FLUX IS NOT AN EXTRAPOLATION OF HIGH
ENERGY QUIESCENT FLUX
IR
X-γ
Israel et al. 2005
• But the IR flux varies with the quiescent flux and/or
with bursting activity
ESSENTIAL SGR PROPERTIES
Radio
Counterpart
NIR
Counterpart
X-ray
Counterpart
SGR1806-20
After giant flare
(transient)
Yes
Yes
Massive star
cluster?
SGR1900+14
After giant flare
(transient)
No
Yes
Massive star Massive Star?
cluster?
SGR0525-66
No
No
Yes
SNR
?
SGR1627-41
No
No
Yes
?
?
SGR0501+45
No
Maybe
Yes
SNR?
?
1E1547-5408
Yes
No
Yes
SNR
?
Host
Progenitor
48M
HOW MANY SGRs ARE THERE?
• 5 or 6 confirmed SGRs (depending on 1E1547)
• Three unconfirmed SGRs: 1801-23, 1808-20, GRB050925
• Some short GRBs could be extragalactic giant magnetar flares
• Muno et al. (2008) estimated <540 in the galaxy, based on Chandra,
XMM data
CONCLUSIONS
• There is good evidence that the known SGRs are magnetars
• There is growing evidence that the members of the magnetar family
(AXPs and SGRs) are very similar to one another
SGRs AND AXPs
OBSERVATIONAL PROPERTIES COMPARED
SGRs
AXPs
Small Bursts
Frequent
Rare
Giant Flares
Yes
No
Quiescent X-rays
Yes, to >100 keV
Yes, to >100 keV
Radio emission
Following giant flares
only
1-2 cases known,
transient
Periods
5–8s
2 – 11 s
Spindown
.5 – 20 × 10-11 s/s
0.05 – 20 × 10-11 s/s
Hosts
Massive star clusters?
SNRs?
•
In 1992, Duncan and Thompson, and Paczyński, independently proposed
that neutron stars with large magnetic fields could explain SGR bursts
and giant flares
•
In 1995, Thompson and Duncan expanded their model to explain the
AXPs
•
Duncan and Thompson called these neutron stars magnetars
•
Motivation:
– High B  low opacity, so L>>LEddington is allowed
– High B  neutron star magnetosphere can contain the energy of the
radiating electrons in a giant flare
– High B causes rapid spindown of newly born pulsar – in the case of
SGR0525, the age of the SNR is 10,000 y, and the period is 8 s
– High B is a reservoir of energy to power the quiescent emission and
the bursts
MAGNETARS
• Definition: a neutron star in which the magnetic field, rather than
rotation, provides the main source of free energy; the decaying
field powers electromagnetic radiation (R. Duncan & C.
Thompson, 1992; C. Thompson & R. Duncan, 1995, 1996)
• Note that the definition does not specify the magnetic field strength
• To explain SGRs and AXPs, however, B must be greater than the
quantum critical value 4.4 x 1013 G, where the energy between
electron Landau levels equals their rest mass
• Some AXPs and SGRs require B~1015 Gauss, so these magnetars
have the strongest cosmic magnetic fields that we know of in the
universe
ORIGIN OF THE MAGNETIC FIELD
•
Unknown, but there are two hypotheses:
1. Fossil field: massive (25 M) progenitor star’s field (104 G or
more) is amplified during core collapse and frozen into highly
conducting compact remnant (the neutron star). Initial period of
neutron star is 4-10 ms, too slow for a dynamo to operate
efficiently.
2. Dynamo amplification: field is generated by convective dynamo in
the proto-neutron star. Initial period is 1-3 ms.
•
These hypotheses are not mutually exclusive
MAKING A MAGNETAR WITH A DYNAMO
(Duncan & Thompson 1992)
• A neutron star undergoes vigorous convection in the first ~30 s
after its formation
• Coupled with rapid rotation (~1 ms period), this makes the
neutron star a likely site for dynamo action
• If the rotation period is less than the convective overturn time,
magnetic field amplification is possible
• In principle, B ~ 3 x 1017 G can be generated (magnetic field
energy should not exceed the binding energy of a neutron star, so
B<5 x 1018 G)
• Differential rotation and magnetic braking quickly slow the period
down to the 5-10 s range
• Magnetic diffusion and dissipation create hot spots on the neutron
star surface, which cause the star to be a quiescent, periodic X-ray
source
• The strong magnetic field stresses the iron surface of the star, to
which it is anchored
• The surface undergoes localized cracking, shaking the field lines
and creating Alfvèn waves, which accelerate electrons to ~100 keV;
they radiate their energy in short (100 ms) bursts with energies
1040 – 1041 erg (magnitude 19.5 crustquake)
• There is enough energy to power bursting activity for 104 y
Thompson & Duncan 1995
B≈1015 G
NEUTRON STAR
•
Localized cracking can’t relieve all the stress, which continues to build
•
Over decades, the built-up stress ruptures the surface of the star
profoundly – a magnitude 23.2 starquake
•
Magnetic field lines annihilate, filling the magnetosphere with MeV
electrons
•
Initial spike in the giant flare is radiation from the entire magnetosphere
(>1014 G required to contain electrons)
•
Periodic component comes from the surface of the neutron star
THE STATISTICS OF SHORT SGR BURSTS ARE
CONSISTENT WITH THE MAGNETAR MODEL
• Burst durations
• Distribution of the time between bursts
• Number-Intensity relation for short bursts
STATISTICS:
DISTRIBUTION OF SHORT BURST DURATIONS
(Gogus et al. 2001)
LOGNORMAL
LOGNORMAL
STATISTICS:
DISTRIBUTION OF THE TIME BETWEEN BURSTS
SGR1900+14
RXTE
Gogus
et al. 1999
SGR1900
Gogus et al. 1999
LOGNORMAL
STATISTICS:
NUMBER-INTENSITY DISTRIBUTION
Götz et al. 2006
POWER LAW
STATISTICS
DISTRIBUTIONS OF SGR PROPERTIES
• Lognormal duration and waiting time distributions, and power
law number-intensity distribution, are consistent with:
• Self-organized criticality (Gogus et al. 2000)
– system (neutron star crust) evolves to a critical state due to a
driving force (magnetic stress)
– slight perturbation can cause a chain reaction of any size,
leading to a short burst of arbitrary size (but not a giant flare)
• A set of independent relaxation systems (Palmer 1999)
– Multiple, independent sites on the neutron star accumulate
energy
– Sudden releases of accumulated energy
OUTLINE
•
•
•
•
•
•
•
•
•
•
•
History
SGRs
1. Bursts
– X-and γ-ray time histories, giant flares, QPO’s
– X-ray afterglows
– energy spectra, lines
2. Quiescent emission in X- and γ-rays
AXPs
Interpretation of the data
Data at other wavelengths: radio, optical
Non-electromagnetic emissions: gravitational radiation
Magnetar locations: SNRs, massive clusters
Magnetar census
Terrestrial effects of giant flares
Extragalactic magnetars
The latest news (SGR0501, AXP 1E1547)
AXPs
• At least 4 AXPs have optical/IR counterparts
• All have an IR excess with respect to an extrapolation of their Xray blackbody spectra
• One or two AXPs display transient, pulsed radio emission
OUTLINE
•
•
•
•
•
•
•
•
•
•
•
History
SGRs
1. Bursts
– X-and γ-ray time histories, giant flares, QPO’s
– X-ray afterglows
– energy spectra, lines
2. Quiescent emission in X- and γ-rays
AXPs
Interpretation of the data
Data at other wavelengths: radio, optical
Non-electromagnetic emissions: gravitational radiation
Magnetar locations: SNRs, massive clusters
Magnetar census
Terrestrial effects of giant flares
Extragalactic magnetars
The latest news (SGR0501, AXP 1E1547)
• Magnetars may be deformed during bursts and especially during
giant flares
• QPOs may be evidence of this deformation
• It follows that they may be sources of gravitational radiation
LIGO LIMITS ON GRAVITATIONAL RADIATION
• Upper limit to GR from GRB070201, an SGR giant flare in M31
(Abbott et al. 2008)
• Search for GR from SGR0501 bursts is in progress
OUTLINE
•
•
•
•
•
•
•
•
•
•
•
History
SGRs
1. Bursts
– X-and γ-ray time histories, giant flares, QPO’s
– X-ray afterglows
– energy spectra, lines
2. Quiescent emission in X- and γ-rays
AXPs
Interpretation of the data
Data at other wavelengths: radio, optical
Non-electromagnetic emissions: gravitational radiation
Magnetar locations: SNRs, massive clusters
Magnetar census
Terrestrial effects of giant flares
Extragalactic magnetars
The latest news (SGR0501, AXP 1E1547)
BUT SGR1900+14 IS NOT ASSOCIATED WITH THE
SNR G42.8+0.6!
If the SGR originated in the SNR,
a proper motion of 110 mas/yr is
implied
SGR1900
DeLuca et al. (2008) have set an
upper limit to the proper motion
using 5 years of Chandra data:
< 70 mas/yr
AXPs
• 3 or 4 AXPs are at the geometrical centers of SNRs
• Implied proper motions are small, and these associations are
considered to be likely
• 1 AXP is in the cluster Westerlund 1
OUTLINE
•
•
•
•
•
•
•
•
•
•
•
History
SGRs
1. Bursts
– X-and γ-ray time histories, giant flares, QPO’s
– X-ray afterglows
– energy spectra, lines
2. Quiescent emission in X- and γ-rays
AXPs
Interpretation of the data
Data at other wavelengths: radio, optical
Non-electromagnetic emissions: gravitational radiation
Magnetar locations: SNRs, massive clusters
Magnetar census
Terrestrial effects of giant flares
Extragalactic magnetars
The latest news (SGR0501, AXP 1E1547)
OUTLINE
•
•
•
•
•
•
•
•
•
•
•
History
SGRs
1. Bursts
– X-and γ-ray time histories, giant flares, QPO’s
– X-ray afterglows
– energy spectra, lines
2. Quiescent emission in X- and γ-rays
AXPs
Interpretation of the data
Data at other wavelengths: radio, optical
Non-electromagnetic emissions: gravitational radiation
Magnetar locations: SNRs, massive clusters
Magnetar census
Terrestrial effects of giant flares
Extragalactic magnetars
The latest news (SGR0501, AXP 1E1547)
GIANT FLARES TURN NIGHT INTO DAY!
Effect of the giant flare of 1998 August 27 from SGR1900+14
Inan et al. 1999
Level of the ionosphere as measured by propagation of
VLF signal from Hawaii (21.4 kHz) descends to daytime
value, due to ionization by 3-10 keV X-rays at 30-90 km
THE GIANT FLARE FROM SGR1806-20
•
The most intense solar or cosmic transient ever observed
•
X- and gamma-ray energy released at the source: 3x1046 erg
•
Measured X- and gamma-ray flux at the top of the atmosphere:
1.4 erg/cm2
•
This should be considered a lower limit to the energy released
(saturation effects, limited energy ranges)
•
It would require ~106 erg/cm2 absorbed by the atmosphere (a
giant flare at 15 pc) to cause an effect similar to “nuclear winter”
(ozone depletion, destruction of the food chain)
•
To cause damage to the biosphere, SGR1806-20 would have to be
about 16000 times closer, at the distance of the nearest stars
•
Probability: 10-6
OUTLINE
•
•
•
•
•
•
•
•
•
•
•
History
SGRs
1. Bursts
– X-and γ-ray time histories, giant flares, QPO’s
– X-ray afterglows
– energy spectra, lines
2. Quiescent emission in X- and γ-rays
AXPs
Interpretation of the data
Data at other wavelengths: radio, optical
Non-electromagnetic emissions: gravitational radiation
Magnetar locations: SNRs, massive clusters
Magnetar census
Terrestrial effects of giant flares
Extragalactic magnetars
The latest news (SGR0501, AXP 1E1547)
ARE SOME SHORT GRBs ACTUALLY MAGNETAR
FLARES IN NEARBY GALAXIES?
•Giant flare begins with ~0.2 s long,
hard spectrum spike
80
COUNTS/0.064 S
60
GIANT FLARE FROM SGR1806-20
RHESSI DATA
100000
40
•Viewed from a large distance, only the
initial spike would be visible
10000
Counts/0.5 s
20
6
7
8
TIME, s
9
•The spike is followed by a pulsating tail
with ~1/1000th of the energy
10
•It would resemble a short GRB
1000
•It could be detected out to 100 Mpc
100
•Some short GRBs are almost certainly
giant magnetar flares, but how many?
10
0
100
200
Time, s
300
400
GRB051103 – A POSSIBLE EXTRAGALACTIC GIANT
MAGNETAR FLARE FROM M81 (3.6 Mpc)
M82
M81
Swift BAT
15-150 keV
(Not imaged)
IPN Error
Ellipse
Eγ=7x1046 erg
Frederiks et al. 2007
GRB070201 – A POSSIBLE EXTRAGALACTIC
MAGNETAR FLARE FROM M31 (780 kpc)
120
KONUS-WIND
GRB070201
2 ms. DATA
COUNTS/2 ms.
80
IPN Error
Box
Mazets et al. 2008
40
M31
0
-400
-200
0
200
TIME, MILLISECONDS
LIGO measurements indicate that this could not have
been a binary merger in M31 (Abbott et al. 2008)
Eγ=1.5x1045 erg
400
5 GIANT FLARE ENERGIES
Assumed distance, kpc
Eγ, erg
SGR1900+14
August 27 1998
15
4x1044
SGR0525-66
March 5 1979
55
7x1044
SGR0044+42 (M31)
February 1 2007
780 (M31)
1.5x1045
SGR1806-20
December 27 2004
8.7
8x1045
SGR0952+69 (M81)
November 3 2005
3600 (M81)
7x1046
OUTLINE
•
•
•
•
•
•
•
•
•
•
•
History
SGRs
1. Bursts
– X-and γ-ray time histories, giant flares, QPO’s
– X-ray afterglows
– energy spectra, lines
2. Quiescent emission in X- and γ-rays
AXPs
Interpretation of the data
Data at other wavelengths: radio, optical
Non-electromagnetic emissions: gravitational radiation
Magnetar locations: SNRs, massive clusters
Magnetar census
Terrestrial effects of giant flares
Extragalactic magnetars
The latest news (SGR0501, AXP 1E1547)
INTEGRAL-IBIS IMAGE OF THE QUIESCENT EMISSION FROM SGR0501
JOINT FIT TO THE QUIESCENT SPECTRUM WITH
XMM AND INTEGRAL DATA (2BB+PL)
N. Rea et al., 2009
PL indices
of the
hard tails:
Rea et al. 2009
SGR 0501+4516
0.95
3.1±0.5
1.5-1.9
Götz et al. 2006
0.73±0.17
1.46± 0.21
0.94±0.16
Different spectral shapes below and above 10 keV ⇒different emission mechanisms
AXP 1E1547-5408: THE INTEGRAL SPI-ACS BURST ZOO
14000
COUNTS/50 ms
1E1547-5408
JANUARY 22 2009
10016 S.
12000
10000
8000
6000
1E1547-5408
JANUARY 22 2009
13838 S.
12000
10000
4000
8000
6000
4000
-8
-4
50000
0
4
TIME, S
8
40000
12
-8
-4
0
4
TIME, S
1E1547-5408
JANUARY 22 2009
19068 S.
30000
20000
10000
0
0
100000
20
TIME, S
80000
COUNTS/50 ms
COUNTS/50 ms
COUNTS/50 ms
14000
40
1E1547-5408
JANUARY 22 2009
24456 S.
60000
40000
20000
0
0
20
40
TIME, S
60
8
12
• 1E1547 was initially classified as an AXP
• But the bursts from 1E1547 are SGR-like
• Two possibilities
– 1E1547 was incorrectly classified; it’s really an SGR
– SGRs and AXPs are really the same type of object
Chandra-INTEGRAL Spectrum of 1E1547
(Courtesy G.L. Israel)
OPEN QUESTIONS
•
What is the number-intensity relation for giant magnetar flares?
•
What is the SGR birth rate and lifetime?
•
What is the progenitor magnetic field?
•
What kind of supernova produces an SGR? Why can’t we detect its
remnant?
•
How are AXPs different from SGRs?
•
Are some high-B radio pulsars actually like magnetars?
•
How is the high energy tail of the quiescent component generated?
BUT…WHAT IF THEY’RE NOT MAGNETARS?
• Fallback accretion disk
• Phase transitions in strange stars
AND LOCATION
•
•
•
•
•
SGR0525-66 lies in the direction of
the N49 SNR in the LMC
This association has been
controversial since March 6 1979
Important to resolve, because
– it is the only unobscured SGR,
and
– if the SGR is in N49, it is the
only SGR with an accurately
known distance and age;
HST measurements reveal no
optical counterpart (Kaplan et al.
2001); accretion disk ruled out
Chandra measurements indicate
that the quiescent X-ray spectrum
resembles that of an AXP,
suggesting that the neutron star
may be intermediate between and
SGR and an AXP
SHOULD YOU BUY A MAGNETAR SHELTER?
(WORST CASE CALCULATION)
• Say each giant flare releases 3x1046 erg
• ~106 erg/cm2 absorbed by the atmosphere (a giant flare at 15 pc)
would cause an effect similar to “nuclear winter” (ozone depletion,
destruction of the food chain)
• Assume magnetars are distributed uniformly throughout the disk
of the galaxy (900 kpc3), and that ~10 are active at any given time
• Probability ~ 10-6
LOCATIONS OF THE FOUR KNOWN SGRS
SGR1627-41
SGR 1806-20
SGR 1900+14
SGR0525-66
N49
LMC
 Magnetars are young objects
MAGNETAR LOCATIONS IN THE GALAXY
MAGNETAR LOCATIONS IN THE GALAXY
M
M
M
M
M
M
M
M
M
M
M
s
M
M
M
YOU ARE HERE
M
MAGNETAR LOCATIONS IN THE GALAXY
M
M
M
MM MM
M
M MM
M
M
M
M
⇒MAGNETARS ARE YOUNG OBJECTS
SGR1806-20 DISTANCE ESTIMATES
8.7 kpc
Bibby et al. 2008
Spectroscopy
Figer et al. 2004
LBV
Eikenberry et al. &
Corbel & Eikenberry 2004
LBV
HI abs.
Cameron et al. 2005
McClure-Griffiths & Gaensler 2005
HI abs.
Corbel et al. 1997
Mol. Clouds
4
8
12
16
DISTANCE TO SGR1806-20, kpc
20
DUST RING AROUND SGR1900+14
Wachter et al. 2008
Spitzer observation
•
If the SGR is at the distance of the
massive cluster, the August 27
1998 giant flare could have created
a dust-free cavity
•
The ring would be illuminated by
the stars in the cluster
SGR1900+14 DISTANCE ESTIMATES
13.5 kpc
Vrba et al. 1999
Supergiant stars
Vasisht et al. 1994
G42.8+0.6
0
4
8
12
DISTANCE TO SGR1900+14, kpc
16
NUMBER-INTENSITY RELATION FOR 5 GIANT SGR
FLARES*
6
N(>E)
4
2
0
44
10
10
45
46
10
47
10
E, erg
*not to be taken too seriously
HOW TO ANSWER THEM
• More detailed theory
• More sensitive instruments, longer observation times
• Another 30 years of data
THE GIANT FLARE FROM SGR1806-20
•
December 27 2004 21:30:26 UT
•
SGR1806-20 was over longitude 146.2º W, latitude +20.4º (near
Hawaii)
•
Detected by at least 24 spacecraft (and probably numerous
military spacecraft) – most of which had no X- or gamma-ray
detectors!
•
The most intense solar or cosmic transient ever observed
•
Measured X- and gamma-ray flux at the top of the atmosphere:
1.4 erg/cm2
•
X- and gamma-ray energy released at the source: 3x1046 erg
•
This should be considered a lower limit to the energy released
(saturation effects, limited energy ranges)
A FEW NUMBERS FOR COMPARISON
• Peak luminosity: 2x1047 erg/s
– Peak luminosity of all the stars in the Galaxy: 8.7x1043
• Total energy released: 3x1046 erg in ~ 1 second
– =300,000 years of the sun’s energy output
– =8x1018 times the yearly world energy consumption
– =8x1019 times the energy in the world’s nuclear arsenal
• But the dose outside the Earth’s atmosphere, 0.14 rem, was
roughly equivalent to a medical X-ray
• To cause damage to the biosphere, SGR1806-20 would have to be
about 16000 times closer, at the distance of the nearest stars
SHOULD YOU BUY A MAGNETAR SHELTER?
(WORST CASE CALCULATION)
• Say each giant flare releases 3x1046 erg
• ~106 erg/cm2 absorbed by the atmosphere (a giant flare at 15 pc)
would cause an effect similar to “nuclear winter” (ozone depletion,
destruction of the food chain)
• Assume magnetars are distributed uniformly throughout the disk
of the galaxy (900 kpc3), and that ~10 are active at any given time
• Probability ~ 10-6
GIANT SGR FLARES ARE SPECTACULAR!
• Occur perhaps every 30 years on a given SGR
• Intense (3x1046 erg at the source, 1 erg/cm2 at Earth), ~5 minute
long bursts of X- and gamma-rays with very hard energy spectra
(up to several MeV at least)
• Are modulated with the neutron star periodicity
• Create transient radio nebulae, dramatic ionospheric disturbances
• Display fast oscillations which provide a clue to the structure of the
neutron star
LESS WELL KNOWN SGR PROPERTIES
Name
SGR1806-20
Possible
Distance, Age,
SNR
kpc
kyr
Association
G10.0-0.3 6 - 15
?
Massive
Star
Cluster?
Yes
SGR1900+14
G42.8+0.6
5 - 15
5 - 20
Yes
SGR0525-66
N49
55
10
No
SGR1627-41
G337.0-0.1 6-11
1-5
No
THE GIANT FLARE FROM SGR1806-20
•
December 27 2004 21:30:26 UT
•
SGR1806-20 was over longitude 146.2º W, latitude +20.4º (near
Hawaii)
•
Detected by at least 24 spacecraft (and probably numerous
military spacecraft) – most of which had no X- or gamma-ray
detectors!
•
The most intense solar or cosmic transient ever observed
•
Measured X- and gamma-ray flux at the top of the atmosphere:
1.4 erg/cm2
•
X- and gamma-ray energy released at the source: 3x1046 erg
•
This should be considered a lower limit to the energy released
(saturation effects, limited energy ranges)
A FEW NUMBERS FOR COMPARISON
• Peak luminosity: 2x1047 erg/s
– Peak luminosity of all the stars in the Galaxy: 8.7x1043
• Total energy released: 3x1046 erg in ~ 1 second
– =300,000 years of the sun’s energy output
– =8x1018 times the yearly world energy consumption
– =8x1019 times the energy in the world’s nuclear arsenal
• But the dose outside the Earth’s atmosphere, 0.14 rem, was
roughly equivalent to a medical X-ray
• To cause damage to the biosphere, SGR1806-20 would have to be
about 16000 times closer, at the distance of the nearest stars
SHOULD YOU BUY A MAGNETAR SHELTER?
(WORST CASE CALCULATION)
• Say each giant flare releases 3x1046 erg
• ~106 erg/cm2 absorbed by the atmosphere (a giant flare at 15 pc)
would cause an effect similar to “nuclear winter” (ozone depletion,
destruction of the food chain)
• Assume magnetars are distributed uniformly throughout the disk
of the galaxy (900 kpc3), and that ~10 are active at any given time
• Probability ~ 10-6
IS SGR1900+14 ASSOCIATED WITH A HISTORICAL
SUPERNOVA? (Wang, Li, and Zhao 2002)
• “During the 3rd month of the 3rd year of Jian-Ping in the period of the
king of Han Ai (4 BC, April), a po star was seen near He-Gu”
• Similar description is found in Korean records
• “po” star had m~5
• In this direction, m=5 corresponds to M=-21; possibly a hypernova
R. Mallozzi, 1998
•
On August 22, 2008, SGR0501+4516 awoke from a long sleep and became
burst-active
•
The first detection was by the Swift BAT
•
This was the first new SGR to be confirmed in 10 years
•
Many ToO observations were called (AGILE, Chandra, INTEGRAL, RXTE,
Suzaku, Swift, XMM)
•
We activated our INTEGRAL AO-6 ToO program on August 27, and observed
the source for about 200 ksec
•
We detected the quiescent emission, and four weak bursts
TIMELINES OF BURSTING ACTIVITY AND TOO OBSERVATIONS
ToO OBSERVATIONS
XTE
CHANDRA
XRT
XRT
XMM
SUZAKU
30
SGR0501+4516
20
10
ep-0
8
3-S
ep-0
8
2- S
ug 08
1-S
ep-0
8
08
ug-
31-A
ug 08
30-A
ug 08
29-A
08
28-A
ug-
ug 08
27-A
26-A
08
ug 08
ug-
25-A
ug 08
24-A
23-A
ug 08
0
22-A
OBSERVED BURSTS PER DAY*
INTEGRAL
*OBSERVATIONS BY KONUS-WIND, RHESSI,
INTEGRAL-IBIS, SWIFT, SUZAKU XIS, AGILE,
AND FERMI GBM
BURST #1
August 27 2008 16:25
COUNTS/50 ms
20-40 keV
40-100 keV
100-300 keV
BURST #2
August 27 2008 22:26
COUNTS/50 ms
20-40 keV
40-100 keV
100-300 keV
BURST #3
August 28 2008 03:32
COUNTS/50 ms
20-40 keV
40-100 keV
100-300 keV
BURST #4
August 28 2008 20:59
COUNTS/50 ms
20-40 keV
40-100 keV
100-300 keV
BURST PARAMETERS
Duration,
ms
Fluence,
erg cm-2
Peak flux,
erg cm-2 s-1 (50 ms)
kT, keV
(OTTB)
August 27 16:25
150
6.6x10-9
4.8x10-8
---
August 27 22:25
200
4.8x10-8
8.4x10-8
---
August 28 03:32
200
3x10-8
1.4x10-7
28
August 28 20:59
50
2x10-8
3x10-7
33
These bursts are very weak; spectra
are typical for SGR bursts
Spectrum of August 28 20:59 burst
OTTB fit
kT=33 keV
SUMMARY
• INTEGRAL-IBIS ToO observation of SGR0501+4516 was successful
• 4 weak bursts were detected. The spectra of two of them were typical
of SGR bursts; the other two were too weak for analysis
• Quiescent emission was detected up to 100 keV; flux was 4x10-11 erg
cm-2 s-1, 18-100 keV, a typical number for SGRs
• XMM/INTEGRAL spectrum indicates a rising νFν spectrum above 10
keV with a hard power law, which is fairly typical of magnetars
INTERMEDIATE
DURATION
BURSTS
(RARE)
“STRANGE”
BURSTS,
1000 TIMES
LONGER
(RARE)
1
1
0
1
0
0
9
0
8
0
U
L
Y
S
S
E
S
S
G
R
1
6
2
7
4
1
J
U
L
Y
2
,
2
0
0
1
2
5
1
5
0
k
e
V
7
0
6
0
COUNTS/32ms
5
0
4
0
3
0
2
0
1
0
0
5
6
7
8
9
1
0
1
1
1
2
1
3
1
4
1
5
T
I
M
E
,
S
E
C
O
N
D
S
INTERMEDIATE BURST ENERGY SPECTRUM:
SUM OF TWO BLACKBODIES, kT=4.3 and 9.8 keV
(SGR1900+14, Olive et al. 2004)
HETE WXM HETE FREGATE
NO GLITCH AT THE TIME OF THE SGR1806-20
GIANT FLARE
HOW MANY ARE THERE?
• When they burst normally in gamma-rays, they are bright enough to
detect anywhere in the galaxy
• When they don’t, you can detect them as quiescent X-ray sources, but
you have to know exactly where to look
• Long periods (decades) with no bursts  possible hidden galactic
SGRs
• Muno et al. (2008) estimated <540 based on Chandra, XMM data