10 Years of Super-Luminous Supernovae: Analytical and Numerical Models

Transcription

1 F.O.E. meeting - 6/2/2015, Raleigh, NC 10 Years of Super-Luminous Supernovae: Analytical and Numerical Models Manos Chatzopoulos Enrico Fermi Postdoctoral Fellow FLASH Center for Computational Science Department of Astronomy & Astrophysics

2 F.O.E. meeting - 6/2/2015, Raleigh, NC Bartender at the Hiberian Irish Pub & Restaurant This day s talk My passport Manos Chatzopoulos Enrico Fermi Postdoctoral Fellow FLASH Center for Computational Science Department of Astronomy & Astrophysics

3 OBSERVATIONAL PROPERTIES SLSN - II Gal-Yam (2012) Gal-Yam (2012) > 50 events discovered (PTF, PanSTARRS, SDSS, RSVP ) Lbol > erg/s trise ~ 10 to > 100 days H-rich and H-poor spectra Some (06tf) show line polarization STUNNING DIVERSITY SLSN - I Gal-Yam (2012)

4 DEMOGRAPHICS HOSTS SLSN II 32% SLSN I 68% SLSN I - Of Ib/Ic subtypes SLSN II - Of IIn/IIL subtypes Rare! (Quimby et al. 2013) SLSN I predominantly found in low-l galaxies: <MV> ~ mag <Mhost> ~ 2 x 10 8 M <SFR> ~ 1 M /yr Lunnan et al. (2014); Leloudas et al. (2015)

5 THE SLSN ENGINE - OUTLINE Ni-56 and Co-56 Magnetar spin-down CSM Interaction Ni-mass SN Ejecta mass SN ejecta mass SN Ejecta mass Initial B-field SN explosion energy SN velocity Initial period CSM mass Opacity/gamma-ray trapping CSM density profile CSM composition SN/CSM separation 3-D CSM geometry Parameter degeneracy, a.k.a. LCs alone are not good enough Chatzopoulos et al. (2013)

6 PISN Scoreboard Event LC Spectra SN 2007bi? N SN 1999as? N CO core masses > 60 M => Γ ad < 4/3 => Contraction => Explosive Burning to Ni-56. Challenges Ni-56/SN ejecta mass, most PISNe intrinsically dim, extreme massloss at observed metallicities, spectra too red compared to observations.

7 PISN + ROTATION Chatzopoulos et al. (2015) Spectra too red compared to o b s e r va t i o n s. M i s m a t c h o f features in contemporaneous epochs (also observed by Dessart et al. 2013). Properties of pre-sn rotation/ metallicity undecipherable in model SN spectra.

8 SN 2007bi - A CONTINUOUS DEBATE Gal-Yam et al. (2009): PISN Based on Light Curve Fits but Ni/Co line blends at late-time spectra. Dessart et al. (2013); Chatzopoulos et al. (2015): Too blue to be true! LC concerns (explosion date) and stellar evolution constraints. Chatzopoulos et al. (2015) Chatzopoulos et al. (2015) MAGNETAR SPIN-DOWN VS H-POOR CSM INTERACTION Dessart et al. (2013b) Chatzopoulos et al. (2014, 2015)

9 MAGNETAR SPIN DOWN Kasen & Bildsten (2010); Woosley (2010); Dessart et al. (2012) One can reproduce SLSN-I like blue/ hot spectra by assuming a constant top-hat energy injection (Dessart et al. 2012). SN ejecta SLSN-I and GRBs have similar hosts (Lunnan et al. 2014). Scoreboard Event LC Spectra SN 2007bi Y Y? Newborn magnetar SN 2008es Y? Challenges SN 2010gx Y? PS1-11ap Y? SN 2012il Y? It s hard to thermalize magnetar radiation in the expanding SN ejecta (Bucciantini et al. 2005). Asymmetrical deposition of energy.

10 CSM INTERACTION Can fit all SLSN LCs (Chatzopoulos et al. 2013). Naturally explains diversity. Consistent with advanced massive stellar evolution/ mass-loss. Chevalier & Fransson (1994) Smith et al. (2010) Scoreboard Event LC Spectra SN 2005ap Y? SN 2006gy Y Y SN 2007bi Y? SN 2008am Y Y PTF 12dam Y? Challenges Hard to compute spectra. Complex conditions for presence/absence of emission features.

11 CSM INTERACTION - LIGHT CURVES Number of fitting parameters can explain diversity but can also give rise to degeneracy (Chatzopoulos et al. 2013). Shells of masses 1 to >10 M needed to explain energetics. -> Episodic mass-loss. CSI models can give hints on massive star evolution a few years prior to explosion (e.g. SN 2009ip). Moriya et al. (2013) [Rad-hydro MGFLD calculation] 100% of SLSN-II almost certainly CSI. L bol [erg/s] 1e+45 1e+44 X Xe3 Xe3m6 Xe3m6r Xe10 Xe10m6 Xm3 Xm6 1e Time [d] Dessart et al. (2015) [Real Rad. Transfer!] Chatzopoulos et al. (2013) [Analytical models]

12 CSM INTERACTION: SLSN-II SPECTRA Only spectral modeling can break the model LC degeneracies between SLSN models and amongst CSI models themselves (Dessart et al. 2013; Chatzopoulos et al. 2013). However this is challenging as it involves radiation hydrodynamics and non-lte radiative transfer. Lagrangian radiative transfer codes designed for monotonically increasing velocity profiles. Dessart et al. (2015) breakthrough: Use Sobolev approximation to calculate interaction-specific (H-lines for SLSN-II) features thereby using the absolute value of velocity gradient to calculate line optical depths. ρ [g cm -3 ] v [cm/s] r [cm] Smith et al. (2010) Dessart et al. (2015)

13 CSM INTERACTION & H-POOR SLSNe Rotationally-induced mixing and episodic massloss (via an LBV or a PPISN mechanism; Chatzopoulos et al. 2013) can lead to total stripping of the H and many times even He envelopes of massive stars. He and C/O-rich shells can thus be ejected forming the CSM environment around compact pre-sn stars. The resulting SN may interact with this H- deficient environment giving rise to some SLSN- I. Gal-Yam et al. (2009) Models of CSM interaction can fit the LCs of SLSN-I (Chatzopoulos et al. 2013a,b; Baklanov et al. 2015). Can some of the features be attributed to CSI interaction? Need for detailed non-lte radiative transfer (Chatzopoulos et al, in prep.).

14 PRE-SN MASS LOSS: SETTING THE STAGE FOR SLSNe The last few days to months of massive stellar evolution are very chaotic and violent. Vigorous pre-sn shell convection (mainly driven by Si but also O shell burning) can generate gravity waves energetic enough to trigger episodic mass loss (Quataert & Shiode 2012). P r e - S N c o n v e c t i o n c a n a l s o significantly change the structure of the progenitor star (Arnett & Meakin 2011). Different ICs for CCSN simulations that can significantly aid susceptibility to explosion (Couch & O Connor 2014; Chatzopoulos et al. 2015; Couch, Chatzopoulos et al. 2015). Chatzopoulos et al. (in prep.) Many supernovae explode inside diverse, complex environments leading to diverse CSI properties.

15 SUMMARY 10 years of SLSN discoveries have yielded a new class of rare explosions that shows striking diversity in terms of LCs and spectra. PISN model unlikely candidate for the vast majority of SLSNe. Spectra too red, constraints from mass-loss at observed metallicities. Magnetar model can fit SLSN-I LCs and perhaps the spectrum of 07bi but uncertainties on thermalization of radiation remain. CSI models can fit LCs of both SLSN-II and SLSN-I and the spectra of SLSN-II. They also naturally explain the observed diversity. Massive stars suffer episodic mass-loss in the days to months prior to SN that can both form a CSM around them but also change the ICs of the stars leading to CC. Need for detailed non-lte radiation transfer for magnetar and H- poor CSI models to determine SLSN-I progenitors.

16 COMMERCIAL BREAK FLASH FLUX-LIMITED DIFFUSION RADIATION TRANSFER WITH RADIATION FLUX LIMITER AWARE HYDRODYNAMICS FOR ASTROPHYSICS Mixed-frame radiation hydrodynamics equations (Krumholz et al. 2007). Updated unsplit hydro solver. Includes variety of flux limiters. Heating by Ni-56/Co-56 decay gamma-ray deposition included. Unified radiation transfer scheme allowing for selective implicit treatment of energy terms. Can do Gray and broad-band LCs of SNe, SN shock break-out, CSM interaction, star-formation Open-source! First release in mid-july (Gray only). Tested on a variety of problems similar to CASTRO paper (Zhang et al. 2011, 2013).

17 COMMERCIAL BREAK FLASH FLUX-LIMITED DIFFUSION RADIATION TRANSFER WITH RADIATION FLUX LIMITER AWARE HYDRODYNAMICS FOR ASTROPHYSICS 2T black-body spectrum Radiative blast wave in the weak and strong radiation/matter coupling limits. Radiation-inhibited Bondi accretion

18 THANK YOU! F.O.E. meeting - 6/2/2015, Raleigh, NC