Imaging of Radio Supernovae

1
Imaging of Radio Supernovae
M. F. Bietenholz
2
Introduction

• Radio emission from a supernova was first
  detected in the 1972 (SN 1970G Gottesman et
  al., Goss et al.)
• First paper about supernova and VLBI was in 1974
  Cass A at meter wavelengths
• First determination of the size of a supernova in
  1983 SN 1979C (Bartel et al.)
• First image of a radio supernova in 1984
  41.95575 in M82 (Wilkinson de Bruyn)

3
41 95595
1984, 1.7 GHz
41.95575, 1984
20 mas
Wilkinson de Bruyn, 1984
4
SN 1986J
1988.8, 8 GHz
1 mas
Bartel et al., 1990
5
Why Image Radio Supernovae?

• Because we can!
• VLBI is the only way to resolve a young
  (extra-galactic) supernova
• To watch the interaction of the expanding
  supernova shell with its surroundings
• To learn about supernovae and the circumstellar
  media of the progenitor stars
• Image the state of the ejecta shortly after SN
  explosion
• Direct distance determination D vmax/?

.
6
What Can We Learn From SN Imaging?

• Interaction of the expanding ejecta with the
  circumstellar medium (CSM) usually the wind of
  the SN progenitor
• Chevalier mini-shell model if you take a
  power-law density distribution for the ejecta and
  the CSM, then
  
• rshock ? tm
• m depends on the ejecta and CSM density profiles,
  and the rate of mass loss of the progenitor
• Stellar wind history of star supernova shock
  front overruns CSM wind with 1000 times faster
  speed
• time machine that records progenitor wind
  history in reverse

7
What Can We Learn From SN Imaging?

• Self-similar model expansion rate of shock
  front,
• rshock ? tm , M/w
• m determined by the density profiles of the
  ejecta and CSM
• ?SN ? r -n and ?CSM ? r -s, m
  (n-3)/(n-s)
• Non self-similar evolution? hydrodynamic
  simulations
• Shell profile absorption in center? Pulsar
  nebula?
• Clumps in ejecta? Clumps in stellar wind
• Rayleigh-Taylor instabilities
• Shock particle acceleration

.
8
10 RSNe Observed with VLBI
Approximately 30 RSNe with flux densities 1 mJy
have been detected (Weiler et al.)
9
Expansion of SN 1986J
? ? t m , m 0.710.11
10
Late Radio Spectrum of SN1986J
11
SN1986J
1988.9, 8 GHz
1999.1, 5 GHz
1990.6, 8 GHz
2002.2, 5 GHz
12
Evolution of SN 1986J
1988 to 2002, 5 and 8 GHz
13
41 31592
1.7 GHz
41.31592 Expansion velocity 2500 km/s
McDonald et al., 2001
14
41 95575
1.7 GHz,
41.95592 Expansion velocity 2000 km/s
Supernova?
McDonald et al., 2001
15
SN1987A
HST difference (false color) ATCA radio
9 GHz (contours) 2000
Chandra X-ray (false color) ATCA radio
9 GHz (contours) 2000
Manchester et al., 2002
16
(No Transcript)
17
Expansion of SN 1987A
Shell model
Shell point sources model

• placeholder

Substantial deceleration!
Manchester et al., 2002
18
SN 1993J in M81
Contours 5 GHz VLA Radio observations of M81
(Nov. 1997) Optical image from A. Sandage, The
Hubble Atlas of Galaxies
19
Astrometry of SN 1993J w.r.t. the Core of M81
Location of explosion center determined to 45 µas
or 160 AU
Peculiar proper motion 18 9 µas / yr 320 160
km/s to south (2 3 of expansion velocity)
20
Expansion of SN1993J
Marcaide et al., 2003
21
Expansion of SN1993J
66 radius measure-ments
22
Deceleration of SN1993J

• Deceleration parameter, m
• ? ? t m(t)

23
Deceleration of SN1993J

• Deceleration parameter, m
• ? ? t m(t)

Hydrodynamical modelling (Mioduszewski,
Dwarkadas, Ball 2001)
24
SN1993J
5.0 GHz
8.4 GHz
17 May 1993, day 50
Radius 520 AU
Opening begins to appear in the West
Opening appears in the North
31 Oct 1994, day 582
25
SN1993J
5.0 GHz
8.4 GHz
23 Dec1994, day 635
Two hot spots appear in the E and W
Third hot spot appears to the South
8 Apr 1996, day 1107
26
SN1993J
5.0 GHz
8.4 GHz
1 Sep 1996, day 1253
Opening to the North begins to fill in
Hot spots shift slightly
15 Nov 1997, day 1693
27
SN1993J
5.0 GHz
8.4 GHz
3 Jun 1998, day 1893
Hotspot to the south-southwest
16 May 1999, day 2271
28
SN1993J
5.0 GHz
8.4 GHz
25 Feb 2000, day 2525
After 8 years, the radius is 12,000 AU
10 Jun 2001, day 2996
29
SN1993J
5.0 GHz
24 May 2002, day 3345 9 years after the
explosion
Expansion is isotropic to 5.5
30
SN1993J
2001.9, 5 GHz
Marcaide et al., 2003
31
SN1993J
8 GHz
Radio Luminosity in center is Nebula
resolution is 8 of the diameter
Composite Image made from 1998 to 2000 data at
8.4 GHz
32
Shell Profile of SN 1993J
Best fit model shell with partial absorption in
the interior
Shell thickness 25 3 of outer radius
8.4 GHz
33
Evolution of SN1993J
8.4 GHz, from t 50d to t 2787d
34
Characteristics of Resolved RSNe
1Marcaide et al., 2002 report strong deceleration
after 1996
35
What Have We Learnt?

• The radio emission comes chiefly from the
  interaction of the ejecta with the circumstellar
  medium, not from a mini-pulsar nebula
• Mini-shell model with power-law evolution
  provides a good first-order description
• At least in the case of SN 1993J (and 1987A)
  there is evidence for more complex structure in
  the ejecta and/or the CSM
• Spherical shells seem common, albeit with
  moderately strong intensity modulation
• Bisymmetric structure, as is expected from
  axisymmetric stellar winds is not prominent
• Protrusions, jets, amorphous remnants.
• Distance determinations

36
The Future of Radio SN Imaging

• Continued observations of current RSNe
• A few RSNe per decade are bright and close enough
  for imaging (before SKA)
• Planned improvements in bandwidth and recording
  speed will increase sensitivity/resolution
• Current global VLBI arrays 2 as sensitive as
  VLA, so improvements in calibration still
  possible
• Phase connection
• Imaging/model-fitting
• Wide field techniques useful for M82, Arp 220.
  We can see and resolve the Crab and Cass A in
  nearby galaxies

37
What can we Expect to Learn?

• Direct distance determination
• Understanding the differences between RSNe types
• Determining the CSM and ejecta density and
  magnetic field profiles (hydrodynamic modelling)
• Stellar wind history, nature of progenitor
• Rayleigh-Taylor instability
• Shock particle acceleration
• Witnessing the birth of a pulsar nebula
• GRBs and SNe at least some GRBs are associated
  with SNe, ie. GRB 980425 was associated with SN
  1998bw, which was also radio luminous.

SN 1993J
38
Supernova VLBI with the SKA

• The practical distance limit for resolving a
  supernova is 30 Mpc (Virgo cluster), where
  22 GHz global VLBI gives you a resolution of
  0.15 mas 3600 AU
• Since 1980, there have been 16 radio supernovae
  with S 22GHz 0.1 mJy and distance
• The phased SKA the VLBA can image a supernova
  of 0.1 mJy at 22 GHz

39
Sensitivity for VLBI

• Notes
• All entries assume 8hours, 2 polarizations, 2
  bit digitization
• ?I is the expected image rms
• Smin is the approximate minimum flux for good
  imaging of a supernova

40
Comparison of RSNe and SNRs
1.7 GHz,
41.95592 Expansion velocity 2000 km/s
McDonald et al., 2001
41
(No Transcript)