1. M. F. Bietenholz Imaging of Radio Supernovae

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: 41.95+575 in M82 (Wilkinson & de Bruyn)

3. 20 mas 41.95+575, 1984 4 1 9 5 + 5 9 5 Wilkinson & de Bruyn, 1984 1984, 1.7 GHz

4. SN1986JSN1986J 1 mas 1988.8, 8 GHz 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 = v max / .

6. 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 r shock  t m 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 What Can We Learn From SN Imaging?

7. Self-similar model: expansion rate of shock front, r shock  t m, 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 What Can We Learn From SN Imaging?.

8. ~10 RSNe Observed with VLBI NameHost galaxy Distance (Mpc) Peak flux (8GHz, mJy) SN1979CM100166 SN1980KNGC6946 62 SN1986JNGC89110100 41.95+575M82 3.21000 43.31+592M82 3.2~3 44.01+596M82 3.2~20 SN1987ALMC 0.0580 SN1993JM81 4100 SN1994IM51 820 SN2001gdNGC5033134 Approximately 30 RSNe with flux densities > 1 mJy have been detected (Weiler et al.)

9. Expansion of SN 1986J θ  t m, m = 0.71±0.11

10. Late Radio Spectrum of SN1986J

11. SN1986JSN1986J 1988.9, 8 GHz 1990.6, 8 GHz 1999.1, 5 GHz 2002.2, 5 GHz

12. Evolution of SN 1986J 1988 to 2002, 5 and 8 GHz

13. 4 1 3 1 + 5 9 2 1.7 GHz 41.31+592: Expansion velocity = 2500 km/s McDonald et al., 2001

14. 4 1 9 5 + 5 7 5 1.7 GHz, 41.95+592 Expansion velocity: 2000 km/s McDonald et al., 2001 Supernova?

15. SN1987ASN1987A 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

17. Expansion of SN 1987A placeholder Shell model Shell + point sources model Substantial deceleration! Manchester et al., 2002

18. Contours: 5 GHz VLA Radio observations of M81 (Nov. 1997) Optical image from A. Sandage, The Hubble Atlas of Galaxies SN 1993J in M81

19. Astrometry of SN 1993J w.r.t. the Core of M81 Peculiar proper motion of SN1993J 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: θ  t m(t)

23. Deceleration of SN1993J Deceleration parameter, m: θ  t m(t) Hydrodynamical modelling (Mioduszewski, Dwarkadas, & Ball 2001)

24. SN1993JSN1993J 8.4 GHz 5.0 GHz 31 Oct 1994, day 582 17 May 1993, day 50 1 mas, 3600 AU Radius 520 AU Opening begins to appear in the West Opening appears in the North

25. SN1993JSN1993J 8.4 GHz 5.0 GHz 8 Apr 1996, day 1107 23 Dec1994, day 635 1 mas, 3600 AU Two hot spots appear in the E and W Third hot spot appears to the South

26. SN1993JSN1993J 8.4 GHz 5.0 GHz 15 Nov 1997, day 1693 1 Sep 1996, day 1253 1 mas, 3600 AU Opening to the North begins to fill in Hot spots shift slightly

27. SN1993JSN1993J 8.4 GHz 5.0 GHz 16 May 1999, day 2271 3 Jun 1998, day 1893 1 mas, 3600 AU Hotspot to the south-southwest

28. SN1993JSN1993J 8.4 GHz 5.0 GHz 10 Jun 2001, day 2996 25 Feb 2000, day 2525 1 mas, 3600 AU After 8 years, the radius is 12,000 AU

29. SN1993JSN1993J 5.0 GHz 24 May 2002, day 3345 9 years after the explosion Position of Explosion Center Expansion is isotropic to 5.5%

30. 2001.9, 5 GHz Marcaide et al., 2003 SN1993JSN1993J

31. SN1993JSN1993J 8 GHz Composite Image made from 1998 to 2000 data at 8.4 GHz resolution is 8% of the diameter Radio Luminosity in center is <30% that of Crab Nebula

32. Shell Profile of SN 1993J Shell thickness: 25 ± 3% of outer radius 8.4 GHz Best fit model: shell with partial absorption in the interior

33. Evolution of SN1993J 8.4 GHz, from t = 50d to t = 2787d

34. Characteristics of Resolved RSNe NameHost galaxy Supernova Type Expansion speed ( current or latest, km/s ) DecelerationMorphology SN1979CM100II12,000 m = 0.95 1 Circular SN1980KNGC6946II<24,000? SN1986JNGC891II6000m ~ 0.7Distorted shell 41.95+575M82?<2000asymmetric 43.31+592M82?9500m > 0.7Circular shell 44.01+596M82??Circular shell SN1987ALMCIIpec3000m ~ 0.1Circular shell SN1993JM81II8500m ~ 0.8Circular shell SN1994IM51Ic<65,000? SN2001gdNGC5033II14,000 to 26,000? 1 Marcaide 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 determinationDirect distance determination Understanding the differences between RSNe typesUnderstanding the differences between RSNe types Determining the CSM and ejecta density and magnetic field profiles (hydrodynamic modelling)Determining the CSM and ejecta density and magnetic field profiles (hydrodynamic modelling) Stellar wind history, nature of progenitorStellar wind history, nature of progenitor Rayleigh-Taylor instabilityRayleigh-Taylor instability Shock particle accelerationShock particle acceleration Witnessing the birth of a pulsar nebulaWitnessing 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.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 < 20 Mpc The phased SKA + the VLBA can image a supernova of ~0.1 mJy at 22 GHz

39. Sensitivity for VLBI Array  ν (MHz) 5GHz  I (μJy) 5GHz S min (mJy) 22GHz  I (μJy) 22GHz S min (mJy) VLBA (current) 32482514070 VLBA + phased SKA 10000.60.301.70.85 Notes All entries assume 8hours, 2 polarizations, 2 bit digitization  I is the expected image rms S min is the approximate minimum flux for good imaging of a supernova

40. 1.7 GHz, 41.95+592 Expansion velocity: 2000 km/s Comparison of RSNe and SNRs McDonald et al., 2001