Transcript Circumstellar interaction in supernovae

Circumstellar interaction in
supernovae
Poonam Chandra
Royal Military Collage of Canada
SUPERNOVAE (SNe)
Massive explosions in the universe
Few months to few years timescale
Energy emitted 1051 ergs (1029 times more than an
atmospheric nuclear explosion)
Shines brighter than the host Galaxy
As much energy in 1 month as sun in ~1 billion years
In universe 8 supernova explosions every second
 Thermonuclear and gravitational collapse
Calcium in our bones
Oxygen we breathe
Iron in our cars
Origin: Massive stars
Nuclear reactions inside a star
4-8 Msun : thermonuclear supernovae
•4-8 Massive star: Burning until Carbon
•Makes Carbon-Oxygen white dwarf
•White Dwarf in binary companion accretes mass
•Mass reaches Chandrashekhar mass
•Core reaches ignition temperature for Carbon
•Merges with the binary, exceed Chandrasekhar mass
•Begins to collapse. Nuclear fusion sets
•Explosion by runaway reaction – Carbon detonation
• Nothing remains at the center
• Energy of 1051 ergs comes out
• Standard candles, geometry of the Universe
Thermonuclear Supernovae
M >8 Msun : core collapse supernovae
• Burns until Iron core is formed at the center
• No more burning
• Gravitational collapse
• First implosion (increasing density and temperature at
the center)
• Core very hard (nuclear matter density)
• Implosion turns into explosion
• Neutron star remnant at the centre.
• Explosion with 1053 ergs energy
• 99% in neutrinos and 1 % in ElectroMagnetic
• Scatter all heavy material required for life
Core Collapse
Supernovae
Based on optical
spectra
Classification
H (Type II)
(Various types-IIn,
IIP, IIb etc.)
No H (Type I)
Si (Type Ia)
No Si (6150Ao)
Thermonuclear
He (Type Ib)
No He (Type Ic)
Specific problems:
Interaction of the ejected material from the
supernovae with their surrounding medium
and study them in multiwavebands.
Shock velocity of typical SNe are ~1000 times the
velocity of the (red supergiant) wind. Hence, SNe
observed few years after explosion can probe the
history of the progenitor star thousands of years back.
105K
107K
Circumstellar
environment
109K
1/R2
SN explosion centre
Photosphere
Outgoing ejecta
Reverse shock shell
Contact discontinuity
Forward shock shell
Chevalier & Fransson, astro-ph/0110060 (2001)
Radio Emission
Radio emission is synchrotron emission due to energetic electrons
in the presence of the high energy magnetic fields.
Radio emission is absorbed either by free-free absorption
from the circumstellar medium or synchrotron self
absorption depending upon the mass loss rate, ejecta velocity
and electron temperature, magnetic field. Both absorption
mechanisms carry relevant information.
Free-free absorption:
absorption by external
medium
Information about mass loss rate.
2
 .

    M uw T


2
ff
3
2
s
R 3
Synchrotron self absorption: absorption by
internal medium
Information about magnetic field and the size.

ssa

2.5
1.5
B
N
rel
X-ray emission from supernovae
Thermal X-rays
versus
Non-thermal X-rays
Date of Explosion :
28 March 1993
SN 1993J
Type : IIb
Parent Galaxy :M81
Distance : 3.63 Mpc
“X-rays from explosion site: 15 years of light curves of SN 1993J”,
P. Chandra, et al. 2008, submitted to ApJ
“Modeling the light curves of SN 1993J”,
T. Nymark, P. Chandra, C. Fransson 2008, accepted for publication in A&A
“Unusual behaviour in the radio spectrum of SN 1993J”,
P. Chandra 2007, AIP Conference Proceedings, Volume 937, pp. 331
“Synchrotron aging and the radio spectrum of SN 1993J”,
P. Chandra, A. Ray, S. Bhatnagar 2004 ApJ Letters 604, 97
“The late time radio emission from SN1993J at meter wavelengths”,
P. Chandra, A. Ray, S. Bhatnagar 2004 ApJ Letters 604, 97
Understanding the physical mechanisms in the forward
shocked shell from observations in low and high frequency
radio bands with the GMRT and the VLA.
Radio emission in a supernova arises due to
synchrotron emission, which arises by the
ACCELERATION OF ELECTRONS
in presence of an
ENHANCED MAGNETIC FIELD.
Giant Meterwave Radio Telescope, India
Very Large Array, USA
On Day 3200…… GMRT+VLA spectrum
Chandra, P. et al.
F
l
u
x
Synchrotron
cooling break at 4
GHz
GMRT
VLA
Frequency
1.5 years later…………. ~Day 3750
Synchrotron
cooling break at
F
l
u
x
~5.5 GHz
GMRT
VLA
Frequency
Synchrotron Aging
Due to the efficient synchrotron
radiation, the electrons, in a
magnetic field, with high energies
are depleted.
.
4
dE
2
e


2
2
2

B sin E


4 7
3m c
 dt  Sync
b
Q(E)E-g
N(E)=kE-g
N(E)
steepening of spectral index from =(g-1)/2 to g/2 i.e. by 0.5
Ecut off
3e
2
 
B
sin

E
3 5
4

m
c
.
E
1
 2
bB t
F
l
u
x
Synchrotron
cooling
GMRT
VLA
Frequency
break at 4
GHz
Synchrotron
cooling break
F
l
u
x
at ~5.5
GHz
GMRT
VLA
Frequency
On day 3200
B=330 mG
On day 3770
B=280 mG
Magnetic Field follows 1/t decline trend
Equipartition
magnetic
field~ 30 mG
Equipartition magnetic field is 10 times
smaller than actual B, hence magnetic energy
density is 4 order of magnitude higher than
relativistic energy density
d break
dt
2
2



R 1/ 2
R 3 / 2
3
1/ 2
1/ 2 
 B0 
t  2t  
t  2t 
 20 
  20 

Diffusion acceleration coefficient
=(5.3 +/- 3.0) x 1024 cm2 s-1
Radio emission in a supernova arises due to
synchrotron emission, which arises by the
ACCELERATION OF ELECTRONS
in presence of an
ENHANCED MAGNETIC FIELD.
On Day 3200…… GMRT+VLA spectrum
Chandra, P. et al.
F
l
u
x
Synchrotron
cooling break at 4
GHz
GMRT
VLA
Frequency
SN 2006X, Patat, Chandra, P. et al. 2007, Science
•In Virgo cluster spiral Galaxy M100
•Feb 4, 2006, 70 million light years away
•Type Ia supernova
SN 2006X, Patat, Chandra, P. et al. 2007, Science
•Type Ia supernova (Thermonuclear supernova)
•True nature of progenitor star system?
•What serves as a companion star?
•How to detect signatures of the binary system?
•Single degenerate or double degenerate
system?
How to investigate?
Search for signatures of the material tranferred to
the accreting white dwarf.
•Narrow emission lines
•Radio emission
•X-ray emission
Till date no detection.
ABSORPTION OF THE RADIATIONS COMING FROM
SUPERNOVA DUE TO THE CIRCUMSTELLAR MEDIUM
SURROUNDING SUPERNOVA.
Observations of SN 2006X:
•Observations with 8.2m VLT on day -2, +14, +61, +121
•Observations with Keck on day +105
•Observations with VLA on day ∼ 400 (Chandra et al.
ATel 2007).
•Observations with VLA on day ∼ 2 (Stockdale, ATel
729, 2006).
•Observations with ChandraXO on day ∼ 10 (Immler,
ATel 751, 2006).
Na I D2
line
Na vs Ca
RESULTS
•Associated with the progenitor system.
•Estimate of Na I ionizing flux: SUV ∼ 5 × 10 50 photons s − 1
• This flux can ionize Na I up to ri ∼ 1018 cm.
•Ionization timescale τi < Recombination timescale τr .
Increase in ionization fraction till maximum light.
Recombination star ts.
• When all Na II recombined, no evolution. Agree with
results.
•Recombination. This implies ne ∼ 10 5 cm − 3
(ONLY PARTIALLY IONIZED HYDROGEN CAN PRODUCE SUCH
HIGH NUMBER DENSITY OF ELECTRONS )
•Confinement: rH ≈ 10 16 cm
Mass estimation
From spectroscopic data:
Na I column density N (Na I) ≈ 1012 cm − 1
log(Na/H)= −6.3.
For complete recombination,
M (H) ≤ 3 × 10−4 ⊙ M.
H-alpha luminosity ~ 1034 erg s-1
From radio:
3 − σ upper limit on flux density F (8.46GHz) <
70 µJy.
Mass loss rate ≤ 10 − 8 ⊙ M year − 1
CSM mass < 10 − 3 ⊙ M Below detection limit.
Nature of the progenitor star
•CSM expansion velocity ∼ 50 − 100 km s − 1 .
•For R ∼ 1016 cm, material ejected ∼ 50 year before!
•Double-degenerate system not possible. Not enough
mass.
•Single degenerate. Favorable.
•Not main sequence stars or compact Helium stars.
•High velocity required.
•Compatible with Early red giant phase stars.
•Possibility of successive novae ejection.
COLLABORATORS
Claes Fransson (Stockholm Obs)
Tanya Nymark (Stockholm Obs)
Roger Chevalier (UVA)
Dale Frail (NRAO)
Alak Ray (TIFR)
Shri Kulkarni (Caltech)
Brad Cenko (Caltech)
Kurt Weiler (NRL)
Christopher Stockdale (Marquette)
…and …. more
SN 1995N in radio and X-ray bands
(Chandra et al 2008, to appear in ApJ;
Chandra, P. et al. 2005, ApJ)
SN 1995N
A type IIn supernova
Discovered on 1995 May 5
Parent Galaxy MCG-02-38-017
(Distance=24 Mpc)
Bremsstrahlung (kT=2.21 keV, NH=2.46 x 1021/cm2. )
Gaussians at 1.03 keV (N=0.34 +/- 0.19 x 10-5) and 0.87 keV
(N=0.36 +/- 0.41 x 10-5)
NeX
NeIX?
Constraining the progenitor mass
LNeX     j NeX ddV
 ne nNeXI 
eff
h NeX
4
Luminosity of Neon X line
ne 
6.77  105
f
Fraction of NeXI to
total Neon
1
2
M Ne  0.016M sun
Cascade
factor
Emissivity of
neon X line
Number density of
neon is ~ 600 cm-3.
Compatible with
15 solar mass
progenitor star
SN 1995N Chandra observations
Total counts
758 counts
Temperature
2.35 keV
Absorption column
Depth
1.5 x 10-21 cm-2
0.1-2.4 keV
Unabsorbed flux 0.6-1.0 x 10-13 erg cm-2 s-1
0.5-7.0 keV
Unabsorbed flux 0.8-1.3 x 10-13 erg cm-2 s-1
Luminosity (0.1-10 keV) 2 x 1040 erg s-1
•How fast ejecta is decelerating?
R~t-0.8
•What is the mass loss rate of the progenitor star?
M/t = 6 x 10-5 Msun yr-1
•Density structure
Density ~ R-8.5
•Density and temperature of the reverse shock
Forward shock: T=2.4 x 108 K, Density=3.3 x 105 cm-3
Reverse shock: T=0.9 x 107 K, Density= 2 x 106 cm-3
COLLABORATORS
Claes Fransson (Stockholm Obs)
Tanya Nymark (Stockholm Obs)
Roger Chevalier (UVA)
Dale Frail (NRAO)
Alak Ray (TIFR)
Shri Kulkarni (Caltech)
Brad Cenko (Caltech)
Kurt Weiler (NRL)
Christopher Stockdale (Marquette)
…and …. more
Synchrotron Aging in SN 1993J
Synchrotron losses
Adiabatic expansion
Diffusive Fermi acceleration
Energy losses due to adiabatic expansion
V
E
 dE 
 E


R
t
 dt  Adia
R
V
Ejecta velocity
Size of the SN
Energy gain due to diffusive Fermi acceleration
E EV
E(R / t)
 dE 





tc
20 
20 
 dt  Fermi
2
4( v1  v 2 )

3v
4 
tc 
v
 1
1 



 v1 v 2 
2
v1 Upstream velocity
v 2Downstream velocity
  Spatial diffusion
coefficient of the test
particles across
ambient magnetic field
Particle velocity
v
E
E

 2 2
2 2
1
dE / dt Total ( R t / 20  ) E  bB E  t E
For
 t
and
B  B0 / t
(Fransson & Bjornsson, 1998,
ApJ, 509, 861)
Break frequency
.
 R
1 / 2
1/ 2 
 break  B 
t
 2t 
 20 

3
0
.
2
Poonam Chandra
2
.
 R
1 / 2
1/ 2 
 break  B 
t
 2t 
 20 

3
0
.
2
2
Acceleration
diffusion constant
Ball & Kirk 1992
For SN 1987A
   21024 cm2 sec-1
(Ball & Kirk, 1992, ApJL)
Scaled value of diffusion coefficient for 1993J
4
 
 2 10 24  2.96 10 24
2.7
cm2 sec-1