The Mysterious Connection between Superluminous Supernovae and Gamma-Ray Bursts STScI Mon, May 23, 2016 4:02 PM Lin Yan: SLSNe-I with Late-time H-alpha emission discovered by PTF total number of SLSNe now over 80 so discovering one more not such a big deal on its own now want extra information new information is coming out now from late-time spectra and detailed features in light curves iPTF13ehe: slow-evolving SLSN-I zt z=0.343 light curve shows similar shape to scaled-up versions of 2007bi, 12dam explosion date is uncertain, though note late-time rebrigtening in r-band early spectra show H-poor CaII, MgII, FeII, SiII ejecta velocity 13,000 km/s why is 13ehe cooler than others at early t = -9 days? much weaker emission at blue end of optical why does 13ehe show no [O I]6300 at late times (+322 days)? unlike SN 2007 bi a pair-instability SN should have massive C/O core where is the H? why don't we see it at early times? - it could be present, but mixed in ejecta, too hot, ionized early too cool, late but that seems unlikely, temp is too low at T = -9 days - it could be in discrete shell recombination timescale is short, less than 65 days so already neutral but then might re-appear when ejecta slams into this H shell toy model: 40 years ago, H envelope is ejected at 300 km/sec mass between 0.03 and 30 solar masses requires 2 x 10^(49) erg to eject it 0 years ago: core explodes optical photons from explosion pass through neutral H shell (would help if we had more early-time spectra) +1 years in future: SN ejecta runs into H-rich shell will produce broad H-alpha line shell is approx 10^(17) cm in diameter is PTF13ehe a rare event? iPTF15esb at z=0.224 is another SLSN-I early spectrum similar to other SLSN-I but +74 day spectrum shows strong broad H-alpha look at pseudo-bolometric light curve of iPTF15esb shows wiggles in light curve over first 50-60 days could these bumps/wiggles be due to multiple shells created in pulsational pair instability pulses? another case: iPTF10aagc limited dataset spectrum at +75 days shows weak broad H-alpha line conclusions - late-time spectra can probe mass-loss activities shortly explosion - detailed light curve features (bumps, wiggles) can constrain models of the explosion - iPTF13ehe data could be explained by pair-instability pulses maybe iPTF15esb also (but a later talk gives another possible explanation) - 10-20 percent of SLSNe-I could have substational H-rich material ejected by episodes before the explosion 4:18 PM Q: is H-alpha blueshifted from rest wavelength? A: yes, shift of -700 km/s. could be due to shock's motion but is a tricky business to measure accurately Q: what about strong iron [Fe II] line at ~5400? Confident of its ID? A: no, not confident. Is a bit of mystery, VERY wide, could be a combination of several features please give me ideas Q: do you really need pair-instability to explain the observed features? A: no, other mechanisms are possible but you need a LOT of mass, which takes a LOT of energy pair-instability could work Q: do you see these sources in X-rays? Collision should produce X-rays A: very good point. tried Swift followup on one source, found nothing limit is 10^(-14) in flux 4:23 PM Alexandra Kozyreva: (Relatively) Fast evolving pair-instability SNe pair-instability SNe (PISN) others have discussed them already ejecta-wind/shell interaction start with VERY massive star, 140-260 solar masses need oxygen core 60-130 solar masses star is supported by radiation so if photons turn into e-/e+ pairs, radiation pressure collapses and star collapses (or undergoes dynamic instability)j explosive O-Si- burning complete disruption, no remnant can it produce enough nickel? need very massive star computational procedure stellar evolution code, one code pair-instability explosion, second code ejecta evolution and light curve, third code = STELLA results: 200 and 250 solar mass progenitors non-rotating, Z=0.001 110 and 127 solar mass final 12 and 34 solar mass of nichel 9 and 2.6 solar mass of He no H light curves compared to observations of PTF12dam PTF12dam falls between these two models fit over period 50 - 300 days after explosion color temperature doesn't look much like models observations show values 10,000 - 20,000 K at 100-150 days with big uncertainties models around 6,000 - 11,000 K at those times overall, GENEC PISNe are good candidates to explain observations comparison between codes radiation-transport codes: STELLA vs. 8 others used 130 solar-mass helium core, 40 solar mass of Ni used 100 solar-mass helium core, 5 solar mass of Ni light curve shapes not so similar now why the differences? nickel-bubble feedback is not important opacity IS important, uncertainties up to 30 percent conclusions: - high-mass PISN may naturally explain some fast SLSNe - no special need for magnetar or ejecta interaction - watch out for opacities 4:39 PM Q: are you related to the famous Russian astronomer? A: no Q: do you expect some helium feature in your models? A: I don't generate spectra Q: what's the problem with our calculations of opacity? do we need more lines, better measurements of existing ones? A: prob not adding more lines 4:40 PM Ken Chen: Cracking the most luminous SNe: multi-dimensional simulations of pair-instability SNe wish to thank EACOA (East Asian Core Observatories Association) history: Woosley's stellar models 80 solar-mass starting point pair-instability model shows many small pulses, no big deal 90 solar-mass star produced big pulse 10^(49) to 10^(50) ergs, could eject outer envelope Woosley models in 2007 matched SN 2006gy pretty well but some 1-D models produced very sharp strong peaks in density not observed but 2-D or 3-D models would not produce such sharp peaks I have done multi-D simulations 80-140 solar mass stars CASTRO code need supercomputers can resolve very small features in turbulent explosions have 2 or 3 pulses in models before the final explosion later shells can run into ones ejected earlier in these small-scale features, ejecta KE converted to optical radiation then tried to move to radiation hydrodynamics 1-D still shows too-sharp peaks in density, temperature 2-D shows very convoluted thin shell 3-D even more complex these multi-D simulations produce very different light curves and spectra one really _must_ use 3-D simulations to get the real behavior 4:51 PM Q: one of your models have 3 pulses. did you evolve it in 3-D, or 1-D? A: time between pulses can be very long, can't follow in 3-D so evolve in 1-D in between Q: what do you choose as point of mapping? A: from the start of pair-instability Q: ? A: most of the ejecta are well mixed, only out to oxygen [??? not sure] Q: how far from spherically symmetry are your models? could we observe asymmetry? A: the 3-D models are bumpy, but still spherical on large scales doesn't destroy the basic spherical symmetry 4:55 PM Alicia Soderbergh: Utility of late-time photometry in disentangling the Zoo of Superluminous Beasts aka "Stellar Autopsies -- the Cold Cases" 3 facets to late-time energy production - radioactivity - engine injection - CSM interaction all these map to observables: colors, decay rate, dust, non-thermal typical supernova lum optical 10^(42) erg/s, due to 56Ni lasts a few weeks can also observe at other wavelengths atypical supernova non-thermal component increases dramatically, starts to overwhelm the optical, thermal emission GRB supernova hard to detect in optical at early times, because totally overwhelmed the non-thermal is winning but late-time photometry can help nearby GRB-SN have late-time light curves consistent with 56Ni decay what about SLSNe? assume peak mag correlates with 56Ni mass, late-time photometry can separate sources of energy SN 2012au, peak mag -18, rebrightened at late times kinetic energy 10^(52) erg spectra at late times show similarities to Ib, Ic, GRB-SN, SLSN even though not very "super" luminous conclusion: luminosity is not foolproof diagnostic for class -- need spectroscopy possible jetted asymmetry? SN 2012au in radio and X-rays detected at both wavelengths ordinary shockwave and CSM properties thus, unusual explosion mechanism must be deep within, not at exterior Don't neglect older events - go back and look again! The Open Supernova Catalog it's great, please use it upload data if you can Conclusions - many flavors of GRB-SN, SLSN - multiwavelength is essential - radio catalogue will be on-line soon - get ready for next SN in MW 5:07 PM Q: last time we made a working group, we got SN 1987A A: yes! Q: is it possible to follow up all the transients we are finding? A: we need first to learn how to interpret the spectra, before we collect too many new spectra the Open Supernova Catalog can help Q: why not look in M31? A: I agree Q: are you prepared to define a class of sub-luminous super-luminous SNe? A: where do you draw the line? Q: mag = -21 A: why that -- does that create super-luminous, super-luminous SNe? Q: it's not important to set limits on luminosity use other parameters to classify them A: I completely agree Q: John Bahcall had a program on HST to do TOO on MW supernova so maybe you should take over as PI even more likely in IR, so JWST is even better