The Mysterious Connection between Superluminous Supernovae and Gamma-Ray Bursts STScI Wed, May 25, 2016 9:03 AM A. MacFayden: Central Engines of Luminous Supernovae shows movie of exploding rocket illustrates shock waves causing medium to radiate important to couple a source of energy to some radiating mechanism Conclusions - long durations possible in collapsars late accretion can occur, plenty of mass remains at late times even 10 days or 100 days after explosion can be important Luminosity goes like (t/10^6 s)^(-5/3) - jet is not same as light just because there's jet, doesn't mean you see SN - dissipation is needed - turbulent cascade may be important - "bounded magnetar" is more powerful "pulsar in cavity", "pulsar in a bottle" - rapid rotation is rare, most NS spins are slow - very asymmetric SNe are rare - squeezed magnetar has high Lum, can have late flares material falling back can squeeze the bubble Collapsar = rotating, collapsing object this "failed" supernova can create a slow jet which escapes can produce some amounts of nickel as well GRB occurs if the jet is ultra-relavitistic but even a slow jet can end up powering a luminous SN "Delayed" SN explosion core collapse creates NS or BH emits neutrinos but massive stars have more massive envelopes, hard to explode getting rapid enough rotation to get 1 ms magnetar is difficult may require binarity or other effects If a) failure of neutrino-powered explosion b) fast-rotating stellar core (j > 3 x 10^(16) cm^2/s) then rapidly rotating BH accretion disk forms, can last for hours/days (?) can create ultra-relativistic jet Surprise! Nickel wind in accretion disk around BH, some energy is dissipated in outer disk some nuclei may escape in a wind, inside of which nuclear reactions happen this wind can end up as 56Ni and emit energy "nickel wind", "afterburner" spinning mag field around central object can exert poloidal pressure on the surrounding material excavates a cavity elongated toward poles If central object has misaligned spin/mag field get "striped mag wind" can cause mag reconnection, which dissipates energy into surroundings how to create nickel? if star "pre-expands" to > 10^8 m, don't make much nickel need to keep small size, high density to create nickel observations of GRB and their afterglows in order to derive proper luminosity, must account correctly for the changing opening angle of the jetted material is easy to OVER-estimate the energy in the jet maybe many GRBs have lower luminosities than previously estimated new simulations can follow jet from center of star, to surface of star, into ISM very important for matching sims to observations good match in some cases Conclusions - long durations possible in collapsars late accretion can occur, plenty of mass remains at late times even 10 days or 100 days after explosion can be important Luminosity goes like (t/10^6 s)^(-5/3) - jet is not same as light just because there's jet, doesn't mean you see SN - dissipation is needed - turbulent cascade may be important - "bounded magnetar" is more powerful "pulsar in cavity", "pulsar in a bottle" - rapid rotation is rare, most NS spins are slow - very asymmetric SNe are rare - squeezed magnetar has high Lum, can have late flares material falling back can squeeze the bubble 9:28 AM Q: nickel wind depends on electron fraction > 0.5 is it possible to get a wind that doesn't create nickel? A: yes, absolutely other regions will create other nuclei Q: what sets the t^(-5/3) scaling of luminosity with time? A: I have a slide on that -- ask me later Q: if outer layers of star are falling onto mag bubble, how does the late-time flaring occur (or escape)? A: perhaps it's when a shell of material with a lumped ang mom falls back onto central engine Q: material from stellar surface remains at head of jet -- then what? A: collision of jet with this "cap" creates reverse shock which is seen at earliest times, rapid drop in light curve then a long coasting phase at const Lorentz factor responsible for plateau in light curve finally the jet decelerates as it sweeps up too much material responsible for final steeper decline in light curve A: don't forget that most stars do not have rapidly rotating cores Q: does fall-back lead to flares? should lead to changes in velocity and in spectrum A: I think the spectrum should be similar 9:36 AM Ken Nomoto: Radiation Hydrodynamics of Circumstellar Interaction to Model the Multi-color Light Curves of Type I Superluminous Supernovae Superluminous Supernovae could be powered by magnetars in some cases ASASSN-15hl and SN 2011kl, perhaps SLSN-I light curves some are wide (SN 2010gx and PTF09cnd), some are narrow SN 2010gx our circumstellar material model can explain it Sorokina et al. 2016 need 10 solar masses of optically thick material ejecta 0.7 solar mass a mass ejection, not a smooth wind light curve model very wide, > 100 days in red bands PTF09cnd -- even wider light curve need 50 solar masses of CSM, and 2 x 10^(51) erg energy injection ejecta 5 solar mass yields light curve model out to over 200 days so by modifying the amount of CSM and ejecta, can explain light curves of varying widths must be careful when applying semi-analytic models to SNe Double peaks in SLSN-I light curves model with CSM, use more compact CSM: 20 solar masses, 1 x 10^(15) cm use two explosions [?not sure I understood this] explosions about 10 days apart ongoing efforts to match observed light curves how can star provide so much hydrogen-poor CSM pulsation of O-burning evolve a 90-solar mass star, find oscillations in Si burning in core conclusions use optically THICK CSM to explain short and wide light curves use repeated explosions to explain double-peaked light curves low metallicity, rotating, massive star circulation causes H -> He -> C+O need to study stability of nuclear burning (espilon mechanism) inside very massive stars if pair-production instability, stars 80 - 140 solar masses if magnetar, star is < 30 solar masses some of these have too little mass for massive CSM 9:50 AM Q: how do you explain deep, strong absorption lines from the CSM in your models? A: I don't know, haven't thought about it Q: how do you get multiple explosions? A: due to oscillations, instability in nuclear burning important in more massive stars 9:51 AM Asaf Pe'er: Poynting Flux Dominated Jets Challenged by Their Photospheric Emission How to explain what we see? Key features broken power law in spectrum, peak at < MeV what radiative processes produce this spectrum? can't be synchrotron FWHM of the spectrum suggests blackbody mechanisms idea: broaden a thermal spectrum geometrical broadening energy dissipation below the photosphere main source of energy is in magnetic field of jet gradually dissipated in the medium converted to KE and heating models predict energy goes in KE of jet -- at least half of dissipated E energy goes into heating -- less than 30% so lum of prompt GRB should be less than luminosity of afterglow but obs show most GRBs this is not true prediction for thermal emission component what sort of spectrum is created? depends on rates of diff photon production processes double Compton, brehstrahhlung, cyclo-synchrotron in magnetized flows, at certain radii, full thermalization impossible but this is still below photosphere, so photons are trapped end up with a spectrum with Wien shape observed Temp should be 3-5 MeV higher the magnetization, lower the thermal flux Prediction: observed spectrum of GRB should be thermal with T = few MeV but observed spectra have temps of < 1 MeV conclusions - thermal component is a natural way to overcome steep slopes in GRBs - highly magnetized outflows predict Lum(prompt) <= Lum(afterglow) - prediction of Wien spectrum with T = few MeV but this is inconsistent with what we see 10:05 AM Q: so if this model isn't right, what is? A: I think mag fields DO play a role in jets I think thermal component does play a role, too 10:06 AM Philip Moesta: Magnetoturbulence in rapidly-rotating core-collapse SNe how do we create the engines at the centers of some SNe? overview: 3D engine dynamics of magnetar explosions how do we form magnetars from simulations to observations core collapse basics need rapid rotation and B-field amplification results in ms-period proto-magnetar in 2D, this works fine what about in 3D? jets don't form so easily MHD kink instability in jet, pretty well known phenomenon outflows do occur eventually, slowly eventually, this mag bubble expands rapidly continuous accretion in the equatorial plane so maybe form a BH, and BH accretion could drive SN later how to form magnetars? possibly MRI + dynamo recently did global 3D MHD turbulence simulations two weeks of real time on supercomputer 10,000 times more expensive than previous simulations results: get toroidal structure around magnetar very strong magnetic field is created but these simulations need to be extended from simulations to observations simulations can do innermost core, t = 0.1 - 1.0 seconds but how to extend to outer, observable regions? idea: parameterize results of central sims, put into full-star model idea: create simplified simulations of ... summary: - 3D explosion dynamics may explain SN-GRB diversity - MRI and large-scale dynamo explain jet formation - we see beginning of magnetar formation 10:19 AM Q: what is the mass of the magnetar you create? A: mass is still rising, too early to say Q: is there anything you can say about extending your results to galactic scales? A: I hestitate to do so, but a lot of the physics is similar Q: relative energies of magnetar and accretion? A: we need to look at this Q: timescales to build up the mag field to magnetar levels? A: working on it Q: did you have initial mag field in simulation? A: dipole field 10^(10) Gauss not enough by itself to be dynamically relevant 10:24 AM Takashi Moriya: Signatures of a Progenitor Binary Companion in Type Ic Superluminous Supernovae iPTF13ehe: slowly declining SLSN Ic needs 15 solar masses 56Ni if powered by Ni late-time detection of H-alpha emission was there an H-rich CSM shell around explosion? shell at maybe 4 x 10^(16) cm other idea: iPTF13ehe was in binary system, companion has H ejecta of SN will strike companion as is modelled in single-degenerate Type Ia explosions some of the companion envelope is stripped, contaminates the ejecta esp the innermost portions of ejecta as SN expands and cools, photosphere shrinks to expose contaminated region theory is the H is excited by gamma-rays from 56Co -> 56Fe (decay time 111 days) not actually observed in SNe Ia could it work in iPTF13ehe? parameters for iPTF13ehe scale the W7 model PISN: ejecta 110 solar, ejecta energy 60x10^(51) erg, 15 solar 56Ni, stripped mass 0.1 solar magnetar: ejecta 35 solar, ejecta energy 40x10^(51) erg, (15) solar 56Ni, stripped mass 0.7 solar CSM: ejecta 5 solar, ejecta energy 10x10^(51) erg, 1 solar 56Ni, stripped mass 140 solar (!) note magnetar model has no 56Ni, but it must supply an amount of energy equiv to 15 solar masses of 56Ni What sort of star would make good companion for this model? need 0.1 solar mass to be stripped follow Liu et al. (2015) if put H-rich MS star of mass 50 solar about 10 stellar radii away, get 0.3 - 0.05 solar masses stripped - good companion mass should be > 20 solar for PISN, > 80 solar for magnetar H-alpha emission features in iPTF13ehe we see blueshift may be viewing angle, could be a redshift in other cases H-alpha lum reduced by 20% from day 270 - 290 expected from stripped model width of 4000 km/s, similar to stripped model predictions Could early brightening be caused by ejecta colliding with companion? depends on viewing angle maybe is too commonly observed to be explained by collision implications binarity can play a key role companion mass in iPTF13ehe of > 20 solar mass, sep < 20 stellar radii conclusions - note this model can yield blueshifted or redshifted H-alpha line (depending on geometry) - if CSM interaction is causing the H-alpha emission then should always be blueshifted, and early H-alpha absorption should be seen also 10:39 AM Q: how important is initial separation of stars? A: factor of two in sep can lead to factor of 8 in stripped Q: will this material be completely photoionized? A: yes, maybe Q: we need to observe systems very early Q: why important in this SN but not others? A: could be a matter of the separation Q: does binarity play a causal role in explosion? A: it can play a role in the _evolution_ of star which explodes Q: magnetar case requires companion to be more massive than star which exploded A: yes, that's unlikely Q: did you run through population synthesis model to predict numbers of the companions which should exist? A: (no time to answer)