The Mysterious Connection between Superluminous Supernovae and Gamma-Ray Bursts STScI Mon, May 23, 2016 9:00 AM Welcome: Andy Fruchter and Ken Sembach Sembach: This should be lots of fun bring together observers and theorists and simulators share your thoughts with each other multiwavelength observations are key in this field, as in many others HST will be working for at least another 5 years, maybe 10 will continue to operate as long as scientifically productive JWST will starting working in 2 years Oct 18, 2008 is launch date all 18 mirrors have been attached to backplane instruments are being mated to backplane now next year, to California, attach to sunshield at STScI, getting ready for JWST new control room will be controlled from this building (as HST used to be) STScI is looking forward to WFIRST should launch in mid-2020s (apparently one can watch this remotely) Fruchter: please try to keep talks on time there is extra time between sessions for discussions 9:08 AM Quimby: Superluminous Supernovae (a primer) Mysterion will help to explain things 1997cy: was most luminous SN up to c. 2000 peak mag -20 type Ia with CSM interaction now called SNIa-CSM GRB detected at same time - coincident? maybe not but now, -20 is not "super" luminous 2006gy: peak abs mag -22 stayed bright for 3 months! integrated light > 10^(51) ergs -- in PHOTONS found by ROTSE-IIIb narrow H, plus broad H, so type IIn (most SNe at this meeting won't have H) 2005ap: peak abs < -22 SLSN-I (no H) spectrum shows wiggles due to O II how to power it? possibly a GRB hidden inside??? late-time spectrum of SNSL-I start with O II, N III wiggles at early time at later times, Fe and Ca lines, similar to Type Ic SNe spectra compared to ordinary SNe at max light, spectral features are weaker in SLSN SLSN much brighter in near-UV, factors of 10-100 why the difference in UV? SLSN stay hot brighter normal SNe cool to 10000 K at 5-10 days SLSN stay hotter than 10000 K for 1-2 months how? pair-instability mechanism could produce large amounts of radioactive ashes example: SN 2007bi late-time light curve has decay lifetime similar to 56Ni but some SLSN do fade more quickly and some too luminous to be powered at peak by 56Ni another possibility: interaction with CSM transfer kinetic E to radiation need high mass-loss rates before explosion narrow emission lines may indicate ejecta/wind interaction how to get big mass loss rate? maybe Pulsational Pair Instability before core can blow apart completely, core contracts but has blown off outer layers can repeat several times yields large heavy winds from outer layers note that those earlier episodes may look like SNe so could resemble a repeating event, years apart example: PTF13ehe big H-alpha line appears in a second brightening another possibility: Central Engine Power central remnant, BH or NS, can emit energy after explosion note GRB produces jets, which go through ejecta that wouldn't heat up ejecta efficiently GRB jets thought to last briefly SLSN last for months X-ray flares could central engine carve out bubble, eventually break out X-rays? SCP 06F6 produced 10^(45) erg/s, many months after initial outburst Double-Peaked SLSN-I often SLSNe have precursor peak often smaller than second, major peak could be related to central engine, magnetar SLSN-I volumetric rate rate may increase at redshift 2-3 by factor of 3-5 up to 400 SNe per Gpc-cubed per year only about 30-100 locally SLSN-I hosts hosts are often faint similar to hosts of GRBs some evidence that host galaxies are similar, other evidence that they are different GRBs have stronger Mg abs line (?) SLSN-GRB spectra optical spectra of some examples are kinda-sorta similar For the future: - what will spectra of first peak look like? - what will spectra of nebular phase look like? - will other GRB SNe look like SLSNe? - how often will X-ray flares be observable? 9:32 AM Q: precursors -- some other SNe do show precursor outbursts A: SNe IIn can have early outbursts a year or so before main explosion Q: has any SLSN been associated with a GRB? A: [waffles] well, maybe, sort of Q: that item with late-time H-alpha -- did it show He in early spectrum? A: no, we didn't see He in PTF13ehe earliest spectrum may be been too late to see He (cooled too much) 9:37 AM Neil Tanvir: Long Duration GRBs Long-duration GRBs have T90 > 2 seconds but this distinction is dependent on detector, redshift dependent, etc. the short-duration GRBs probably due to merger of compact binaries GRB 970228 was first GRB with optical afterglow later obs showed host galaxy at z = 0.65, small, star-forming showed that GRBs were cosmological light curve showed late-time bump in red GRB 980425 / SN 1998bw first direction connection GRB with SN Type Ic with broad lines, expansion velocities > 20,000 km/s Long GRB light curves rapid variability in prompt emission -- compact progen? emission produced via dissipation after ultra-relativistic expansion? (non-thermal spectrum) requires little pollution by baryons "Relativistic Fireball" standard picture gamma = 300 jet produces prompt emission via internal shocks afterglow via shocking of ambient medium but why so much variation from one GRB to the next? GRB Hosts mostly active star-forming, low-luminosity, low-ish metallicity generally trace regions of star formation progen must be massive stars, age < 10 Myr? Collapsars stripped hydrogen envelope allows easy jet escape favored at low metallicity rapid rotation helps too GRB 030329 = SN 2003dh first SN associated with "high luminosity" GRB optical spectrum very similar to that of SN 1998bw GRB 130427A = SN 2013dq high GRB luminosity 10^(54) erg, but low redshift most other GRB/SN events < 10^(51) erg two populations of GRBs? low-luminosity < 10^(48)ish and high-luminosity up to 10^(52)ish low-luminosity bursts show differences in prompt emission low-lum have shock breakout, not internal shocks? but then, why are SNe in both types so similar? most SNe assoc with GRB look alike light curves look similarish may show common decline-rate/lum relationship Supernova-less GRBs could they be short GRBs? GRB 060614 shows short spike, then extended period of emission other short GRBs show big short spike, then lower-level extended is it possible that some collapsars fail to explode? The engines X-ray light curves show complex behavior plateaux, steep declines, flaring if due to late accretion, why so lumpy? some GRBs show small (tiny) precursors, approx 100 sec earlier hard to explain them with most models The environments most GRBs occur in low-metallicity galaxies ultra-long GRBs recent discovery of ultra-long-duration GRBs last about 10,000 seconds, rather than 100-300 seconds example: GRB 111209A need more observations! Final thoughts - what channels produce long GRBs? single, binary, both, neither? - need to be rare, but enough to explain observed rate - must produce diversity seen in prompt and afterglow properties - relationship to other GRB classes - relationship of GRB-SN to other classes of exotic SNe 10:00 AM Q: will a jet always punch through the atmosphere of its star? A: if it gets through the star, yes, it should keep going but perhaps if CSM is really dense, jet might be stopped ... Q: what is the best number for lowst possible observed rate density? A: I haven't looked at the numbers lately ... a few tens per cubic-Gpc per year let me get back to you later Q: host galaxies of low-lum GRBs look same as hosts of others should we look for other parameters? A: people are studing local host galaxies intensively but all samples are small, alas most long-GRBs seen at high redshift, expect evolution to low redshift I am not sure what you should look for 10:05 AM Brian Metzger: Transients associated with Magnetar Birth magnetar = rapidly rotating strongly magnetized neutron star millisecond "magnetars" - how do they get their strong fields? perhaps rapid rotation at birth in principal, 10^(52) ergs available in rotational KE could amplify mag field to 10^(17) gauss magneto-rotational instability alpha-omega dynamo rotation period (much) less than convective overturn time magnetar should eventually spin down spin-down luminosity is of order 10^(47) erg/s spin-down times shorter for high luminosity, ranges hours -> weeks GRB has strong jets need spin-down time minutes-hours, lum 10^(49)-(51) erg/s SLSN maybe spin-down time days-months, lum 10^(43)-(45) erg/s how to make a bright SN via magnetar shell of gas expanding, decreasing in density key: make spin-down time approx same as diffusion time (and weeks-to-months to agree with obs) toy models mag fields 10^(13) - 10^(14) Gauss look best for SLSNe initial B = 10^(14) gauss with rotation rate of a few millisec if spin too fast or too slow, can't power SN maybe normal GRBs have very strong B-field 10^(16) and very short initial rotation, decays quickly Ultra-long LGRBs could they be slightly weaker B-field, so spin-down longer? GRB 11120A is possible example of magnetar-powered SN assoc with long GRB Kulkarni: A magnetar can't power both a GRB and a SLSN [he's not here to defend himself] but Crab Nebula sorta looks like a possible counterexample toy model of magnetar-powered SLSN magnetar drives strong wind into ejecta wind is anisotropic, can create jets 3-D models are hard, still working on models hybrid LGRB-SLSN by varying B-field and initial rotation rate, can produce a range of events with diff luminosity and gamma-ray burst durations low-lum goes with short GR duration, high-lum goes with long GR duration Tidal Disruption Events shows some similarities to jet-driven SNe Extreme event: ASASSN-15lh radiated energy 10^(52) ergs -- wow! almost at the limit of rotational energy of magnetar helps to have a massive NS but then light curve had re-brightening in UV what caused that? can we look into the heart of the burst? in optical, wait for a few months to see into center in UV/X-rays, opacity much higher must wait for years, decades but can inner UV/X-rays ionize the ejecta, make it transparent? idea: could see turn-on in UV/X-rays weeks/months after optical? shows model for evolution of millisecond pulsar wind nebula some similarities to accretion disk around BH one case, break out in 1-3 months to some X-rays very sensitive to ejecta mass: small mass, easier to break out super-luminous X-rays from SLSN SCP06F6 shows X-rays at age = 70 days very luminous, 10^(45) erg/s what about UV rebrightening of ASASSN-15lh? can't explain as UV breakout, maybe ionization of shell changes opacity from bound-free to free-free Conclusions: [list of 10 items, too many to type] 10:32 AM Q: what is connection of these theoretical magnetars with observed magnetars? we have 28 observed magnetars in MW, for example A: there are a lot of differences probably most MW magnetars don't form this way so there must be multiple ways to form magnetars Q: B-field above 10^(15) Gauss are too high A: simulations show small-scale amplification to such high values Q: this ionization breakout model, would that produce spectral features in late-time spectra? A: yes, should produced more highly-ionized species Q: [didn't get it] Fruchter: poster room has big screens -- see demonstration please send poster electronically to organizers