Day 7416: 2 ,303 Years of SN1987A

1. O Day 7416: 2 ,303 Years of SN1987A •  ADS: ~2826 (~2.7/week) refereed papers (since 1987) (in 45 minutes???) How old were you when this issue appeared? Patrice Bouchet CEA/DSM/Sap Saclay

2. Heating releases a few nucleons  e-+ p →e + n Escaping e carry off energy and lepton number 5 x 109 K 1011g/cm3 • escatter through charged + neutral current (mean free path < ~1m at the center!): •  Initially the SW has enough energy .. BUT: • as it travels, it heats and melts Fe (loss of ~8 MeV/nucleon) • neutrino emission is accelerated  another energy loss 3x1014g/cm3 t = 10ms Summed losses ≡ initial energy carried by SW r The energy released by further collapse is trapped GM2/2R ~ 2.5 1053 ergs (M=1.4Mo; R=10km)

3. DELAYED MECHANISMS Mezzacappa et al., 1995, 2000 Burrows et al., 1995 • t~0.01s  the shock stalls at R~200-300km (~equilibrium) • t >0.5s  neutrino heating of the nucleon “soup” left in the wake of the shock revives the SW (T~2 MeV + very high entropy): Energy is IN THE RADIATION, NOT IN THE MATTER • The pressure exerted by this gas pushes the shock outward: 99% of the 1053 ergs released in the collapse is radiated in υof all flavours • t(leak out) ~3s M M Shock Gain (Heating) Region Ln Convective/SASI Unstable Ln Shock Shock Ln Ln Cooling Region Gain radius Shock M M

4. Chronology • Collapse  Core bounce  proto-NS  NS ~1 s ~10 ms ~0.5 s • g/cm3: 1010  1011  1012  1013  1014 ~100 ms ~25 ms ~5 ms ~1 ms • t ~ 1 min  shock crossed He (R~ 500 x 103 km) • Reverse shock forms • t ~ 3h: shock emerges: flash UV  ionisation • Envelope expands  cools (14000 K to 6000K)

5. Neutrinos • Temperature: ~ 4±1 MeV • Decay time: ~ 4.5 s • Just about right for neutron star formation; results close to modern theory • Most of them νe • Fluence at Earth 5.0±2.5 x 109 cm-2 • Core radius = 30±20 km • Total νeEnergy: ~ 4±1 x 1052 ergs • TotalνEnergy: ~ 3±1 x 1053 ergs • MBaryon = 1.45±0.15 Mo; • MGravit.= 1.35±0.15 Mo ¯ Mont Blanc (LSD) ~4.7 hours earlier? 2-stage explosion in a rapidly rotating collapsar could explain the difference between LSD/IMB-KII υdetections(Imshennik & Ryazhskaya, 2003) ¯ ~1 nanogram of νthrough IMB and KII and only 1 in 1015 were captured  ~500 grams through the entire Earth ≡ 15MegaT of TNT (1 million people experienced 1 SN1987A νe event in their body and ~300 experienced 2 events)

6. Light Echoes • R310, R430; W700, S730, N980; R1170 complex (5 echoes); SE3140, N3240 • 3-dimensional structure (Xu et al., 1995). l.yr Distant echoes: interstellar clouds (P. Tisserand: ~1200 real EROS2 images from july 1996 to feb. 2002) Rings are not matter but a geometrical effect 10 Ly Nearby echoes: Napoleon’s Hat etc.  few solar masses ejected by progenitor

7. The Progenitor Star ? Sk -69°202: B3 Ia; Teff = 16300 K; R = 46.8Ro • Why blue giant, not red giant? • Low metallicity(Shklovskii, 1984; Arnett, 1987; Hillebrandt et al., 1987): Ni rich shell? • Mass loss(Maeder & Lequeux, 1982; Maeder, 1987): light curve and slow V ejecta? • Blue Loops (Summa & Chiosi, 1970): must have been RSG (Woosley, 1988) • Mixing: 56Ni up to ~3000 kms-1, H down to ~500 kms-1? • Convective mixing induced by rotation(Weiss et al., 1988) • semi-convection at low abundance of heavy elements(Woosley et al., 1988) • evolutionary effect in a close binary system(Podsiadlowski & Joss, 1989) NTT / Wampler et al., 1990  new constraints! (at least rotational effects & convective mixing) • Menvelope = 18±1.5 Mo • MHe = 6 ± 1 Mo, MH,Envelope = 5-10 Mo • MFe = 1.45 ± 0.15 Mo • MNS = 1.40 ± 0.15 Mo (2-3 1053 ergs) • Heavy Elements ejected = 1.5 ± 0.5 Mo • (≤1500kms-1) Mpre-SN = (19.4 ± 1.7) Mo

8. First Observations • R-T instabilities: mix the C, O, He layers when the outgoing shock front passes through, but cannot account for the high velocity of Ni (~3100km/s) (Kifonidis et al., 2000), nor the observed polarization. • Convection: size and scale too small to explain asymmetries + NO EFFECT ON NEUTRINO HEATING and shock evolution(Bruenn & Mezzacappa, 1994) • Rapid & differential rotation of progenitor(Steinmetz & Höflich, 1992): explains polarization because of blue SG • Rotation of the collapsing core: introduces axial symmetry but weak effect on the explosion (Zwerger & Müller, 1997). • Magnetic fields + rotation 2 high-density supersonic jets from the collapsed core  too strong B and jets must stop after <3s(Khokhlov &Höflich, 2000) • Precursor too massive for a “good” explosion mechanism • No spherical symmetry (polarimetry) • Strong mixing (Bochum event)

9. Utrobin, 2007 • Single star models: • How massive? (Utrobin 2005) • Rotation tends to suppress the blue solution by increasing the He core mass, but seems necessary to break spherical symmetry prior to the explosion (Woosley et al., 1997) • LBV? (Smith, 2007) • Binary star?(Podsiadlowski, 1992, Morris & Podsiadlowski, 2006)

10. HYDRODYNAMIC MODEL Woosley, 1997 Utrobin, 2007 VNiMin= 900km s-1 E=1.0 x 1051 erg No 56Ni VNiMax= 2500km s-1 VNiMax= 4000km s-1 ii E=1.5 x 1051 erg VNiMin= 960km s-1 No 56Ni VNiMax= 2500km s-1 ii Utrobin & Chugai, 2005 Need of both photometry and spectroscopy

11. Light Curve Evolution • Shock through the surface, 3.84h after neutrinos: T~3x105 K  UV flash  ~3h, R X 10  V ~ 6.4 • As envelope expands it flows through a recombination front: Radiation doesn’t diffuse to photosphere but photosphere moves to radiation;energy is released BY recombination but comes from the SHOCK • After H, He recombination releases energy (shock, recombination itself, and radioactivity that had diffused out while “awaiting” the recombination front.) • Radioactive energy deposition comes from Compton scattering of γ-ray lines (56Co 847, 1238 keV) Recombination wave Radioactive tail 56Co Dust Formation Release of trapped radiation Radioactive decay SAAO Envelope transparent (in the optical) Freeze-out phase + Radioactive tail ESO, CTIO Expanding Photosphere iii iii Ring emission He rec. Radioactive tail 44Ti Ejecta emission H rec. Snuc = M(56Ni) x [3.9 x 1010 e-t/τ(Ni) + 7.2 x 109 (e-t/τ(Co) – e-t/τ(Ni)) erg g-1 s-1

12. FREEZE-OUT The recombination & cooling time scales become comparable with the expansion time scale: conversion of energy in radiation is not instantaneous any longer emitted luminosity remains greater than instantaneous radioactive power deposition(Fransson & Kozma, 1993) UVOIR BLC 56Co / 57Co = 2 X = 2 x 1037 erg s-1 P = 5 x 1037 erg s-1 57Co Total Input 44Ti Observed Bolom. luminosity e+ i 22Na 56Co Bouchet et al., 1996

13. The Dust IAUC4746 March 1, 1989 • Clumps • Silicates? Lucy, Danziger, Gouiffes & Bouchet, 1989, 1991 Bouchet & Danziger, 1993

14. Ejecta emission = “Hot” dust • Dust still present at day 6067 (Bouchet et al., 2004) • And at day 7241 (Bouchet et al., 2006, 2007) • 90 K < TDust,Ejecta < 100 K • MDust,Ejecta = 0.1-2 x 10-3 Mo • LIR = (1.5±0.5) x 1036 ergs-1 N: 10.36μm (Δ=5.30μm) T-ReCS/Gemini-South Fν = 0.3±0.1 mJy! Fν = 0.45±0.05 mJy! Flux increase due to heating by Reverse Shock? Oct. 20,2003 Dec. 26, 2006

15. Inner Debris HST, Jan. 2007 - SAINTS • Glowing: 44Ti decay • Interior dust clouds • Cold! < 300 K • Stirred, not blended • Fe bubbles: ~1% of mass, ~50% of interior volume Axisymmetric ejecta:Wang et al., 2002 Radioactive elements at t=250s 56Ni He, O, Ca Synthesized in progenitor Red Shifted : Blue Shifted Nonrelativistic Jets-induced explosion Why don’t we see a compact object? • Optical, near IR: obscured by black cloud? • X-rays: < cooling neutron star. Debris may be opaque at 1 keV. • Absorbed luminosity should emerge as far IR.

16. Radius: R ~ 0.6 lt yr Expanding: V ~ 10 km s-1 Density ~ 3 x 103 – 3 x 104 cm-3 Glowing mass ~ 0.1 MSun Nitrogen-rich Circumstellar Structure (Michael et al. 2003) • “Standard” Model • RSG  outer envelope BSG • Gas expulsed by slow RSG wind (550kms-1) concentrated into dense equatorial plane • High-velocity low-density isotropic BSG wind for final ~20 000 yr • Faster BSG wind overtook RSG wind • BSG photoionizes and heats gaz from RSG wind (Chevalier & Dwarkadas 1996) Martin & Arnett, 1995

17. Triple Ring system Why three rings? Rotation needed for the equatorial plane, and RSG are too big  Podsiadlowski BUT Woosley, Chevalier, Dwarkadas, etc… • Single rotating star:hydrodynamic formation due to ionization and heating of the cool RSG wind(Meyer, 1997, 1999) • Binary system:impulsive mass loss from primary star, formation of a thin dense shell, and the expansion of 2 jets(Soker, 2002) • Binary mergers:mass loss from a rotationally distorted envelope following rapid in-spiral of a companion inside a common envelope(Podsiadlowski, 1992; Morris & Podsiadlowski, 2006) • LBV: unstable LBV eject and shape their nebulae when BSG (Smith, 2007)

18. PN: MYCN18 (HST/WFPC2) N. Smith LBV: HD168625 Morris & Podsiadlowski, 2006 (Document STSCI)

19. The Reverse Shock McKee, 1974: Expanding debris are decelerated by the CSM which causes a shock to propagate inward through the SN material (+ Chevalier, 1982) • High velocity debris cross the RS at v~ 12 x 103 kms-1 • Freely streaming H atoms in the RS ~ 8000 kms-1 • Post-shock ions = 2000 kms-1 •  Fast atoms & Slow ions Michael et al., 2003 Extinction by dust in the ring Collisional excitation of neutral H from the debris crossing the RS  radiative shock seen as very broad, high-vel Lyα & Hα emission Lyα↑ Resonant scattering: Lyα↓ • No cylindrical symmetry Heng et al., 2007; Smith et al., 2005 Emission from the Reverse Shock

20. The “Bleach Out” of the Reverse Shock • At t=18 yr, Lαfrom RS~ 15Lo≈ 2.3 x 10-3 Mo yr-1(x 4 since 1997) • Lα continuum from gas shocked by the forward blast wave ionizes neutral H in the debris before they reach the RS: •  when the inward flux of ionizing photons exceeds the flux of H approaching the RS Preionization shuts off the RS emission Smith et al., 2005

21. 2007 Hotspots! Challis, 2007 2006 – 2003 difference Pun et al., 2002 • Ring has brightened by factor ~ 3 • Hotspots still unresolved • Have not fully merged • Are the last spots in denser regions than the ones where the first spots appeared? • What caused fingers? • Why so regularly spaced? 1994 Unfolding the ring! (Garnevich, 2006)

22. ACIS Images 2000–2007 Ring-like Asymmetric intensity Developments of X-ray spots becoming acomplete ring as the blast wave crosses the inner ring Surface brightness increases  Now ~x18 brighter than 2000 Lx (0.5-2keV) = 2.1x1036 ergs/s No point source at center N Park, 2007 E 1 arcsec

23. X-ray Imaging N E ACIS (1999-10): Burrows et al. 2000 1991 ROSAT/HRI (5” pixels) HEASARC/SkyView Green-Blue: ACIS Red: HST Contour: ATCA 1 arcsecond Park, 2007

24. Prompt Phase • Burst of emission seen by MOST on day 2; peaked on day 4 (Turtle et al. 1987) • Power law decay, faded by day 150 • Synchrotron in BSG wind (ρ r–2)(Storey & Manchester 1987; Chevalier & Fransson 1987) RADIO EMISSION Ball et al. 1995 • Late Phase(Gaensler, 2007) • after ~3 yr: impact with dense RSG wind • α~ -0.9  optically thin synchrotron emission, steep electron density and small compression ratio in the shock (vs. ~ -0.5) • Consistent multi-wavelength picture of reverse-shock emitting region: interaction is with dense gas in equatorial plane • Radio-source same size as optical ring Manchester et al., 2002

25. Limb brightened • Bright lobes to • east and west • Eastern lobe • brighter than • western lobe, • & brightening • faster •  Same as X-rays Radio Imaging Gaensler, 2007 ATCA 9 GHz super-resolved (0.5 arcsec) ATCA 9 GHz diffraction limited (0.9 arcsec)

26. X-Ray Light Curves 0.5-2 keV 3-10 keV Chandra (0.5 – 2 keV) d ~ 6200 ATCA X-ray Flux (10-13 ergs/cm2/s) X-ray Flux (10-13 ergs/cm2/s) ROSAT Chandra (3 – 10 keV) Chandra/ACIS ROSAT (Hasinger et al. 1996) Similar rates of hard X-ray and radio The same origin for them? due to softening of X-ray spectrum? Forward shock enters a “wall”? X-ray (2005-7) vs. Optical (2005-4) X-rays: (Park et al., 2004, 2006) Radio: Gaensler & Staveley-Smith Image: ACIS 3-8 keV Contours: ATCA 9GHz Image: ACIS 0.4-0.5 keV Image: ACIS 0.5-2 keV Contours: HST (Peter Challis)

27. PHYSICAL INTERPRETATION • Park et al., 2002, 2004; Zhekov et al., 2005  2-shocks model • At t = 18 yr: • Soft X-Ray = Decelerated, slow (300-1700 kms-1); kT=0.51keV; ne~6300 cm-3 • Hard X-Ray = High-speed (3700±900 kms-1); kT= 2.7 keV; ne ~ 280 cm-3 Eli Michael/Colorado Low speed oblique radiative shock: optical/UV Slower shock in high-density knot: soft X-rays High speed shock: radio, hard X-ray

28. MID-IR Emission Bouchet et al., 2006 11.7μm 18.3μm T-ReCS (day 6526) Graphite Carbon Silicates Temperature Optical Depth Spitzer (day 6190) Thermal emission from shock-heated silicate dust T6067 = (180±15) K M6067= (0.1-1) x 10-5 Mo T6526 = (160±15) K M6526= (3 ±1) x 10-6 Mo

29. Dust Heating Mechanism T-ReCS/HST ACIS/11.7 mm ACIS/18.3 mm ATNF/11.7 mm ATNF/18.3 mm • Where is the dust? • In the X-ray emitting gas? • In the denser UVO emitting knots? • What heats the dust? • Collisional heating? • Radiative heating? IR-to-X ray flux ratio IRX ≈ 1! (Tgas≈ 2x107 K  IRX ≈ 100) • Dust severely depleted in the shocked gas: • Grain destruction by the SN shock wave? • Inefficient production in the progenitor? (Dwek, 1987) SPITZER, 2007:

30. ATCA/ Chandra /HST (day 6300) HST/11.7μm (day 6526) HST/18.6μm (day 6526)

31. Physical Picture 11.7μm(Bouchet et al., 2006)and HST(Challis, 2006) EQUATORIAL RING McCray, 2007 HOT FINGERS Optical/Soft X-rays IR?? SHOCK WAVE ? HOT GAS NS/BH REVERSE SW Hard X-rays COOL EJECTA Radio Cf. Michael et al. 1998 SAINTS (F250W)

32. SN 1987A at 20 Years • Explosion mechanism and progenitor?? • HST images: optical spots dominate entire inner ring; blast wave crosses ER • Inner ring detected in the mid-IR at day 6067: shock-heated silicate dust • Dust still present in the ejecta at day 7241: heated by RS?  Reverse shock approaches the central debris (?) • Soft X-ray, mid-IR& radioimages resemble optical image: origin mid-IR? • Ratio [hard (> 3keV) X-rays/radio] ~ Cst. • Soft (~0.5 – 2 keV) X-rays increased rapidly after hotspots appeared; • dominated by the decelerated shock since day ~6000; l.c. makes a turn-up at • day ~6200 as ring mid-IRfluxat day ~6000 • - X-rayradial expansion ratereduces since day ~6200, shock velocity reduces to 1400 km/s; radio expansion constant at ~ 4700 km/s (Rradio = Roptical) Hα=red; OIII=green; F250W=blue (Bouchet et al. 2006). HST - SAINTS mid-IR vs HST (2005-1)

33. Why No Pulsar Detection NOW? • Possible that neutron star has accreted matter and turned into a black hole?: 56Co which powered the l.c. shows that very little mass could have fallen back + BH truncates a gradually decreasing flux of neutrinos + doesn’t produce bursts(Woosley, 1988) • Maybe not beamed toward us - if slowish pulsar, expected beaming fraction ~ 0.2 (Manchester, 2006) • Possiblyobscured by dust • Pulsar magnetic field may take time to develop • A slow, low E pulsar would not pulse at optical or X-ray wavelengths (except maybe thermal emission from NS surface) • Although outer parts of nebula probably have low optical depth, we really know very little about conditions right in centre- could be absorption/scattering of radio pulses . Keep searching for a radio pulsar and point X-ray source!

34. “New” Pulsar Detection ? Middleditch et al., 2000: • Several detections during 1992 – 1996 at different frequencies • Power faded after 1993; last detected 1996 • Emission with a complex period modulation near 2.14 ms • Frequency of the signals followed a consistent and predictable spin-down [~(2-3) x 10-10 Hz s-1] over the several year • Modulation of the 2.14 ms period with a ~1000 s period, which complicates its detection • Precession due to deformation or crustal density distribution not symmetric about the axis of rotation Santostasi, Johnson, & Frank, 2003: • possible asymmetric deformation that causes the precession • main mechanism for the loss of rotational energy due to emission of gravitational radiation •  Continuous source of gravitational wave detectable with LIGO II in a few days (106 years for LIGO I): 2013

35. Forecasting SN1987A NACO 2007 • 2017 celebration: • X-ray, optical ring: ~ 10 x brighter than today • Hotspots will merge • Reverse shock emission will vanish • Interior debris will begin to brighten • Circumstellar matter will begin to glow • Spectacular images of NT radio emission from ALMA • Compact Object: JWST?, ALMA? • Gravitational waves from LIGO II? ( based on McCray, 2007)

36. Forecasting SN1987A • 2027 celebration: • X-rays, optical: ~ 100 x brighter than today • Will clearly see interior debris and circumstellar matter • Newly synthesized elements will begin to cross reverse shock • + ???? ARE WE REALLY MADE FROM THE ASHES OF STELLAR DEATH?

37. How/Why Could a Star Explode? • Direct Hydrodynamic:always fails? (works for <15Mo + soft EOS) • Nuclear-burning aided?(Mezzacappa et al. 2006) • Neutrino-Driven Wind, spherical 1D(Burrows 1987):Lowest-mass massive stars (O/Ne/Mg cores; 8.8 Mo,Kitaura et al. 2006) • Convective/SASI (Spherical Accretion Shock Instability) Neutrino-Driven Wind 2D (Blondin et al. 2003) might workfor 11.2 Mo of WHW2002,(Buras et al. 2006): other progenitor problematic • Acoustic Power Mechanism:after delay, all progenitors explode(Burrows et al. 2006,2007); High entropies favor r-process • Magnetic fields, neutrino mixing and realistic neutrino transport (Mezzacappa et al. 2006) • MHD Jet Explosions:Rapid rotation necessary(e.g., Burrows et al. 2007b) works for Hypernovae and GRB The Key feature of almost all mechanisms is the Breaking of Spherical Symmetry  3D

38. Radial Expansion X-Rays 4700 ± 100 kms-1 Radio t<5000d: 3600km/s t>5000d: 4700km/s ~ 35 000 kms-1 Gaensler et al., 2007 Racusin et al. 2007 rapid deceleration in X whilst ~ constant in radio + Discrepancy in radius between radio and X-rays  ???

39. MAGNETIC FIELD  Radio and hard X-rays come from relatively low density gas between blast wave and reverse shock How (where) are relativistic electrons accelerated? Non linear kinetic theory of Cosmic Ray: shock modification and strong magnetic field amplification in ALL the young SNR(Berezhko, 2005; Berezhko & Ksenofontov, 2000, 2006 ) • Radioemission spectrum • Considerable synchrotron cooling of high energy e- which reduces their X-ray synchrotron flux • Expected γ-ray energy flux at TeV-energies ~ 2 x 10-13 erg cm-2s-1 G-SNRs source population of the G-CR

40. Limits on Properties of a Central Pulsar • No evidence for central source (PWN or pulsar) at any wavelength: optical luminosity limit ~8 x 1033 erg s-1(Graves et al. 2005), X-ray limit (2-10 keV) ~5 x 1034 erg s-1(Shtykovskiy et al. 2005) • Radio limit of central source from 8 GHz image ~ 1 mJy • Assume flat spectrum  20 GHz, LPWN ~ 3 x1031 erg s-1 • For the most conservative limit on E of central pulsar, assume PWN only emits at radio frequencies and LPWN = EPSR • For Po = 200 ms, EPSR = 3 x1031 erg s-1, then Bo ~ 6 x 1010 G . . . Manchester, 2007 Well within the range of possible pulsar birth parameters PWN limits do not rule out a perfectly plausible 20-year old pulsar at the centre of SN 1987A