Radiative-transfer Modeling of Supernova Spectra and Light Curves

Application to Massive-star Explosions

Luc Dessart1 and D. John Hillier2 1 :Laboratoire d’Astrophysique de Marseille, France

2 :University of Pittsburgh, USA

In collaboration with Roni Waldman & Eli Livne (Racah Inst., Jerusalem)

General Aims:Infer SN ejecta properties Radiative Transfer

Connect to Progenitor properties Stellar EvolutionConnect to Explosion mechanisms Grav. Collapse, Explosive burning etc.

Synopsis

• General context and relevance

• Observational Background: SN typing, SN diversity

• Theory Background: Stellar evolution, Explosion mechanisms

• Radiative-transfer modeling: Method, processes, approximations,

dependencies, e.g. line blanketing & opacity, electron-scattering, non-LTE,

time-dependence, decay energy & non-thermal processes

• Application to SN light-curves and spectra: II-P, II-pec, IIb, Ib, Ic

Core-collapse Supernovae (CCSNe): Key phenomena of modern astrophysics

• Associated with massive-star death (M>8M collapse)

• Progenitors of stellar-mass compact objects: SN connection (II,I, dark SNe)?

• Explosion mechanism ? Theoretical challenge. Upper mass cut? Robustness?

• Extreme /T => Neutrino and GRW emitters (quest for a galactic SN).

• Light-curve/spectral diversity heterogeneous progenitor/explosion properties

• Luminous phenomena: Probes of distant Universe, cosmology (Ia)

• Connection to Long-duration GRBs & highly-energetic Type Ic CCSN explosion.

• Connection to observed super-luminous SNe, faint transients, interacting SNe?

• Prime targets of current/future deep blind surveys for transients: PTF, Pan-

STARRS, Sky-Mapper, LSST (2015+)

SN typing: Based purely on spectral signatures

Type I SN: absence of hydrogen lines

Ia: no-HI, no-HeI, SiII 6300A feature

Thermonuclear runaway in envelope of accreting Chandrasekhar-mass white-dwarf in binary system

Ib: no-HI, HeI

Ic: no-HI, no-HeI

Fe-core collapse of an evolved (H-deficient) massive star => Wolf-Rayet

Type II SN: presence of hydrogen lines

Fe-core collapse of an initially less massive star => BSG-RSG

=> SN typing reflects diversity of progenitor properties

Finer classification of CCSNe

• II-Plateau: RSG progenitor (most common)• II-pec: BSG progenitor (rare; e.g.: 87A)• II-L: Mysterious set. Incudes the most luminous SNe known.

Light-curve morphology

Line-profile morphology

Spectral evolution

• IIn, Ibn: Narrow HI/HeI lines

• IIb, i.e. II Ib

• Ic/GRB: SN with -ray emission

Connection to GRB

Smith et al. (2011)

Diversity of Light curves

Car during outburst

Diversity of Spectra

Variations in composition, ionization, excitation, T, V(m)

Multiple paths to massive-star death.

Wide range of properties at death due to:

M,dM/dt,Z, (single) + mass transfer (binaries)

• Envelope stripping: wind mass loss and/or mass transfer. Key for SN diversity

• Single star: Mdot is f(M,L,Z) => RSG/BSG/WNh/WN/WC/WO II-P/II-pec - IIb – Ib – Ic

• Rotation influences transition between RSG & WR at death, i.e., II & Ibc SNe

• Binary stars: Mdot via mass transfer! « WR » from 10-25 M progenitors

• dM/dt, DL/dt = f(Z) => Type II vs. Type I SNe, Ib vs. Ic, GRB/SN, hypernovae without GRBs?

• Puzzle: LBVs as progenitors of CCSNe?

=> HUGE MESS!

(Courtesy of Doug Leonard)

Core evolution to gravitational collapse and explosion

• Stellar evolution: progression toward increasing c and Tc ; Binding energy ladder: H He C O

Ne Mg Si Fe => degenerate core

• Electron capture (producing neutrinos) + photo-dissociation => Pressure deficit

• Collapse when M>Mch=5.83 (Ye)2 M on dynamical timescale (10-100 ms)

• Collapse halted at nuclear densities (~2x1014g/cm3), bounce, shock formation

• Fe white dwarf -> p,n,e-,e+, neutron star: Gain of 1053 erg of grav. energy.

• Energy losses at core bounce (photo-dissociation + neutrino losses): Shock wave stalls

• Explosion requires shock revival:

• Neutrino-driven explosion:

• Magneto-rotational explosion: MRI. Fast rotating progenitor Fe core

• Energy liberated ~ 0.1-10B; Explosive nucleosynthesis: ~0.001-0.1M of 56Ni

Diversity of CCSNe: reflects complexities of stellar evolution (envelope & core properties) & variations

in explosion mechanism/properties

Chronology of events in the life of a CCSN

• 1 sec: Core collapse, bounce, shock revival

• 1 min to 1 day: shock propagates and breaks out (1st EM signature). Fallback? NS vs. BH formation?

• At breakout: Erad~Ekin ; Erad>>Eth ; cont ~106

• Mins to days: Final ejecta acceleration to homology (VR)

• Ejecta properties: Ekin~1051erg, Mejecta~ few M, Vexp~3000km/s, M(56Ni) ~ 0.1M

• Generic subsequent Evolution controlled by

Cooling (Expansion & Radiative losses)

versus Heating (Radioactive decay & Recombination).

modulo Transport (dynamic radiative diffusion --- opacity/composition/ionization, dT/dr!)

Their variations cause the diversity of CCSN Light Curves and Spectra

• Weeks to months: Photospheric phase (1)

• After a (few) month(s): Transition to Nebular phase (1)

• 1-10n years: SNR, CSM interaction, light echoes

Radiative-transfer conditions: Before shock emergence

Shock emergence (shock breakout)

tSBO ~ 70s (R/R) (vs/104km/s)-1

tSBO~ 1min (WR), few mins (BSG), ~1d (RSG)

vshock~104km/s

Post-shock region: vr~3000km/s, T~106K, ~ const.Radiation-dominated shock (P~aT4)Radiation trapped behind shock

R★~1-103R

1D radiation-hydrodynamics simulations (Dessart et al. 2010)

Rarefaction

Before explosion: Big M -> Big ~ M/R2 ~ 106-11 -> Big tdiff ~ 3 R / c ~ 103-5yr.

At large tshock << tdiff : Photons are trapped in the flow

Diffusion wave initiated when tshock ~ tdiff or c /3vshock ~ 10

Radiative Precursor / Burst @1010-1011L

Duration: function of H R =>1-103s

High-energy radiation: X-ray (Ib/c; e.g., SN 2008D) or far-UV (II-pec,II-P)

106 10 1 Atmosphere/Wind

Mean free path: p << R p ~ R

Rphot

tshock ~ tdifftshock << tdiff

tshock >> tdiff

Envelope

Eddington factor: fEdd=1/3 fEdd~1

Radiative-transfer conditions: Shock emergence

Radiative-transfer conditions: After Shock emergence

1.2B RSG explosionDessart et al. (2010)

At breakout: Erad ~ Ekin

Within minutes to days: dPrad/dr Etot ~ Ekin

homologous expansion: RSN >> R★; v(m) = const.; (m) = 0(m)(texpl/t)3

Subsequent evolution controled by cooling, heating & transport

SN radiation influenced by cooling

• Cooling through expansion primarily

• dE=dQ-PdV ; Prad >> Pgas: E=aT4V, P=E/3

• => if dQ=0 then dT/T=-1/3 dV/V. Since dV/V=3dR/R => T ≈ 1/R

• Explosion of a WD: R0=108cm, RSN=1015cm => RSN/R0=107

• => T drops from 109K to room T in ~2 weeks!

Velocity [km/s] Velocity [km/s]

SN radiation influenced by heating

1. Recombination energy: e.g. 13.6eV per HI (weak).

2. Radioactive decay energy: 56Ni 56Co 56Fe (-rays positrons). Key for Type I Sne!

56Ni 56Co : 1.75MeV per decay, half-life=6.07d

56Co 56Fe : 3.74MeV per decay, half-life=77.22d

3. Exceptional circumstances: Magnetar spin-down (Eth), interaction (Ekin Eth)

Monte-Carlo code to follow -rays trough ejecta subject to Compton Scattering and photoelectric absorption

>>1: assume local deposition as heat. <1: non-local, non-thermal

Treatment of -rays:

SN radiation influenced by transport

• Fdiff ~ -dB/d ; B is the Planck function (B~T4)

Radiative Diffusion driven by dT/dR and inhibited by opacity

R/l. Random walk requires steps so tdiff = l/c = R/c

• Note: For the sun, tdiff ~105years, <l>~0.1cm

• tdiff is a characteristic time scale for energy leakage from SN ejecta

• tdiff : function of expansion rate, mass, scale, opacity, ionisation etc.

• Diffusion ≠ Trapping! No diffusion if dT/d but trapping is >1

Supernova Radiative-transfer

Phot. (~1015cm, 8kK, 5000km/s)Cont. Radiation escapes (=2/3)

T LTEEdep, DJ/Dt, DF/dt

Outer boundaryFree streaming

Thermalization depth:J ~ B

Homology:V ~ R

Non-LTE Time-Dependent Radiative Transfer Modeling with CMFGENCo-moving frame formulation for homogously-expanding ejecta (MKH75)

Rate Equation:

Energy Equation:

where

RTE 0th moment:

RTE 1st moment:

Rad

iati

on

Gas

Co

up

lin

g

+ Dedecay/Dt

(Excitation + Ionization)

& charge conservation

1-D Non-LTE time-dependence using CMFGEN(Hillier & Miller 1998; Dessart & Hillier 2005,2008, 2010,2011ab; Hillier & Dessart 2012)

Simultaneously solves the radiative-transfer equation in the CMF frame, the statistical-equilibrium and

the energy equations; Fully implicit solver; partial linearization of SEE.

Accurate description of I/J/H: moments of RTE with all important terms in v/c, /t, /,/, /r

RTE solved for at ~105 frequency points (with coupling). Coverage: ~10A to ~5m

Detailed description of the gas: 25 species & 15 ionization stages. Non-LTE D/Dt ionization

Large model atom: few 10000 levels and few 100000 transitions.

Approximation: Use of Super-Levels. Easy check on approx by switching to full levels.

Non-LTE: All important radiative + collisional rates included explicitly.

Non-LTE line blanketing: All continuum and line opacity sources included explicitly

Solution of Spencer-Fano equation for non-thermal heating/ionization/excitation rates

Adaptive grid: ~7 pts per -decade at each time (asset over hydro: mass grid); [10-8 to 106 ]

SN II-P SN1987A

1-D Non-LTE time-dependence using CMFGEN(Hillier & Miller 1998; Dessart & Hillier 2005,2008, 2010ab)

Physical consistency: stellar-evolution + hydro input: Xi(m), T(m), (m), v(m) r(m)

=> use SN light to constrain pre-SN evolution and explosion

Full-ejecta simulation, e.g. no “artificial” boundary conditions, Xi stratification

Decay energy: Computed with Monte Carlo -ray transport code (local or non-local)

Model requires ~5-10Gb, i.e. (NT=2000) x (ND=100) x (NBNDS+1=4) x 8

Time step: t = 0.1t => 45-50 steps to go from 0.3 to 21d, or 10 to 1000d => 3 months

Application to Supernovae1) Steady-state mode

2) Time-dependent mode

Non-LTE Steady State ApproachApplication: Modeling of Young Type II SNe

The case of the Type II-P SN 2005cs in NGC 5194 (Dessart et al. 2008)

Steady-state models good at early times to constrain the Reddening, composition, expansion rate, distance/radius (EPM)

Study Dependencies!

• Z modulates metal-line blanketing

(v) modulates line and cont

• H/He in SN II-P: influences UV

flux but HI lines poorly sensitive

Degenerate effect of E(B-V) and T. Use lines

Spatial grid

Use optical-depth scale, not mass scale (e.g. hydro code)

Good resolution of photosphere; Eddington factors 1/3 1

Naturally adjusts to resolve recombination fronts

Converged results with ~100 depth points.

HI

HeI

Dessart & Hillier (2010)

Model atom --- Line Blanketing

• Many species: H, He, CNO, IME, IGE• Many ions (I-VII) treated simultaneously • Huge model atom to account for all species/ions and

sources of opacities (lines and continuum).• Sources of atomic data: Opacity project etc. Essential.• In non-LTE, need for all rates (radiative & collisional)• Strong line blanketing

Model atom --- Line Blanketing

Difficult to obtain converged results, especially in UV, UB bands Accuracy of atomic data? Illustration for importance of FeI, NiII and larger FeII model (big huge atom)

Effects associated with Electron-scattering:Frequency redistribution

• photon scattering with free electrons causes frequency redistribution

• Non-coherent scattering in CMF caused by the thermal motion of scatterers: V thermal

• Coherent scattering in CMF due to expansion, Redshift in Observer’s frame: Vexpansion

• Vexpansion > Vthermal the redshift dominates: P-Cygni profile with enhanced red-wing flux

• Vexpansion < Vthermal non-coherent redshift/blueshift dominates: Symmetric profile (SNe IIn)

Effect of varying Vphot on H morphology

Electron-scattering: Flux ‘’Dilution’’

• Scattering-dominated atmosphere => << => = • Eddington Approximation and dB/d =const. => Flux 2√ B(1/ √3 ) -> 1/ √3 : thermalization depth

-> 2√ : Factor of “dilution” (<<1)

• Introduce of ‘’dilution’’ factor Fobs = (Rphot/D2 B(TC) ; (Rphot -> Rphot)

Gezari et al. (2008)Example with early post-breakout models:

SEDs fitted with B(Tth) and ~0.6

Mihalas book; Dessart & Hillier (2005b)

Non-LTE effects

Solution of the statistical-equilibrium equations using all collisional/radiative rates. About 2000 levels/variables => Yields non-LTE level populations, ionization, line-blanketing

High density => Strong collision & weak scattering => LTE Low density => weak collision: Dominance of scattering => Drives gas out of LTE Large departure coefficients in regions of 1 but we recover LTE at depth!

LTENon-LTE

Dessart & Hillier (2011)

non-LTE effects: HeI lines in young SNe II

SN87A obs

LTEModels

Flu

x

Eastman & Kirshner (1989)

Dessart & Hillier (2005): non-LTE model for SN87A Dessart et al. (2008)non-LTE model for 05cs

non-LTE: Account of radiative rates in the

determinations of level populations

non-LTE allows good fit to HeI lines using a

‘’standard’’ BSG/RSG He abundance at early times

non-LTE key for abundance determinations

Spectral Modelling: time-dependent effects

Utrobin & Chugai (2005)

• High V, low trec ~ R/V time-dependent effects (UC05, DH08)

• Retain D/Dt terms in SSE and Energy equations

• Key to yield strong recombination lines in the absence of Lyman/Balmer flux

• H strength at recombination epoch sensitive to Ne rather than profile

D/Dt: No

D/Dt: Yes

Dessart & Hillier (2008)

-ray escape and the nebular-phase decline rate

• Fneb supports full -ray trapping in SNe II-P and II-pec up to few 100d => Higher mass

• Fneb suggests -ray escape as early as ~50d in SNe IIb/Ib/Ic => Lower mass

Dessart et al. (2011)

Dessart & Hillier 2010,2011; Dessart et al. 2011ab

Non-LTE time-dependent Radiative transferFull-Ejecta Simulations

1. SN II-pec (87A): evolution from 0.3 to 21d of BSG-progenitor model

lm18a7Ad (Woosley, priv. comm.)

2. SN II-P: evolution from 10 to 1000d of 1.2B explosion of 15 and 25M

RSG progenitors.

3. SNe IIb/Ib/Ic: evolution from 1 to 100d of 1.2B explosion of (quasi)

hydrogen-less cores produced from binary-star evolution (Yoon et al.

2010).

Light-curve evolution of SNe II-P, II-pec, Ib: Radiative-transfer simulations on full ejecta using physical input models

• Rapid fading after shock breakout

• Post-breakout plateau: LPlateau function of R

• Possible re-brightening from 56Ni/56Co decay. Delay function of mixing.

• High-brightness phase function of M (large ), R (cooling), M56Ni (heating)

• Transition to nebular phase when ~1; Lnebular function of M56Ni and -ray trapping

• Not considered: binary (Kasen 2010) or CSM interaction, Magnetar radiation (Maeda et al. 07)

Confrontation of simulations to observations => M*, R*, Eexpl, 56Ni

LCs are degenerate/ambiguous: confusion Ia/b/c, difference M & Mejecta etc.

(DH10,DH11a; Dessart et al. 2011)

Inputs

pre-SN Star

M

[M]

R

[R]

Mejecta

[M]

Eexpl

[B]

M56Ni

[M]

SN II-P RSG 15(single)

830 10.9 1.2 0.08

SN II-pec (87A)

BSG 18(single)

47 15.6 1.2 0.08

SN Ib WN 25(binary)

10 3.6 1.2 0.24

Comparison to observations of SN1987A

Agreement at 10% level except in the blue (opacity issue)Supports 18M BSG progenitor, R~50R, and Eexpl~1.2B

Dessart & Hillier (2010a)

Observation Simulation

Comparison with SN1987A spectra

Simulations of SNe II-P: 15 and 25M RSG progenitor stars

Dessart & Hillier (2011)

Neb

ular

pha

seP

hoto

sphe

ric p

hase

LP ~ 2-3 x larger than <Lobs>~3x108L=> progenitors R-M too large?

Spectral Evolution

Photospheric Composition vs. Time:Probe from surface to core!

Comparison to SN 1999em (II-P) spectra

300d

40d

70d

[OI]

H

[CaII]

• General agreement at all times: supports RSG

progenitor

• Specific disagreement with H at nebular times: neglect

of non-thermal processes

• [OI]: emission from core + envelope; Key CCSN feature!

SN II-P ejecta kinematics: Vphot vs. [OI] width

• Smartt, van Dyk and Co. find SNe II-P stem from MRSG of 8-17M

• Use nebular [OI] to constrain the main-sequence mass MMS

• MMS Mhe , but dM/dt => Same final mass but bigger core!

=> Higher-mass progenitors eject oxygen at a larger velocities

Observations: v(OI) < 1500km/s => MMS < 20M

Dessart et a.l. (2011)

Study of SNe IIb/Ib/Ic.

What distinguishes them from Type II CCSNe?

Main-sequence mass?

Rotation?

Metallicity?

Explosion mechanism etc.?

Wolf-Rayet-star Properties

• Single WR stars have long been proposed as progenitors of SNe Ibc

• Properties of WR stars: 106L, Mdot~10-5M/yr, v∞=103km/s, M≥10M.

• Emission line spectra => Easy detection out to few Mpcs! (WR search by Crowther/Bibby)

• Final masses ≥10M for most WR, higher for WN, and even higher for WNh?

Crowther (2007)

WR Masses in binaries

Single WRs as progenitors of SNe IIb/Ib/Ic?

• CCSN Rates: 48% of II-P, but 14% of Ic, 7% of Ib, and 10.6% of IIb (Smith et al.

2010) => Not an IMF reflection but calls for a binary component

• Non-detection of any Ibc progenitor conflicts with single luminous WR stars

Problems with SN IIb/Ib/Ic observations:

1)difficulty to produce low-mass cores at death for Type I CCSNe.

2) difficult to produce He-rich envelopes (with residual H) for Ib (IIb).

(Low-mass) Binary channel

• Binary-star evolution naturally produces low-mass stars with degenerate Fe cores from

10-25M main-sequence stars (Yoon et al. 2010). Favored by IMF.

• Progenitor properties: low L, weak winds, no emission lines => NOT genuine WR stars!

• He up to 50% of the total mass.

• Low-binding energy and similar core as 15M RSG => common exposion mech.?

• Brighter and more massive companion when primary explodes. Naturally explains the

existence of SNe Ibc in high luminosity region, even for low-mass binaries.

He

O

56Ni

Observed SN IIb/Ib/Ic Light curves

Drout et al. (2010)

• Nearly always discovered on the rise (suspicious).

• Rise time to peak of ~20 days --- Same as a SN Ia!

• Narrow peak (20d).

• Faster decline than for full -ray trapping (not shown)

• SNe IIb, Ib, and Ic have similar LC properties.

• Scatter in peak brightness. Similar magnitude of scatter for SNe II-P.

SN IIb/Ib/Ic LC - Simulations

• Ensman & Woosley (1988) using artificially stripped 15-25M progenitors explode hydrogen-less cores and reproduce the basic morphology of SNe Ib LC - no discussion of spectra though.

• Dessart et al. (2011,2012): Simulations of both spectra and light curves based on (realistic) binary-star progenitors. Prediction of post-breakout plateau, early peak, narrow peak, fast decline at nebular times. Some agreement with obs.

=> SNe IIb/Ib/Ic have LOW-MASS ejecta!

EW88

The case of SNe IIb

• SNe IIb and Ib have similar LCs. Diff. is HI lines• Our work: HI easy to reproduce even with very little

hydrogen (=> Large rate of IIb?)

=> Low-mass ejecta with little H excludes known WNh stars => Supports close low-mass binaries

• Binary-star progenitor model

• M=18M, Mejecta= 2.91M, M(H) = 0.0007M,

• X(H) = 0.1 at 50000 km/s

• X(H) < 0.001 at 10000 km/s

• M(56Ni)=0.18M (unmixed)

H

Dessart et al. (2011)

SN Ib: Position of the « problem »

• SNe Ib: Strong HeI lines at 10000km/s from a few days to peak and beyond

• Issue: HeI ionization/excitation potential really high => difficult to excite levels or

photoionize/recombine at a cool photosphere.

• Non-thermal processes are key (Lucy 1991).

• Origin: radioactive decay -> Compton scattering -> high-energy (non-thermal) electrons.

Subdominant in number compared to thermal electrons with Maxwellian distribution at

local T of a few 1000K but key non-thermal ionization/excitation effects.

• Magnitude of non-thermal processes conditioned by distribution of -ray energy

deposition.

• Problem: 56Ni produced at the highest and T at base of ejecta in the first second.

Velocity initial ~ 1000 km/s BUT HeI lines seen with 10000km/s width. => Non-local edep.

But gamma-ray mean free path short for several weeks throughout most of the ejecta.

Mixing is essential.

• Large scale asymmetry of explosion: propels 56Ni along 1-2 directions.

• Small scale asymmetry: RT but weak in Type I SNe. Intrinsic structure of SN shock with 56Ni fingers etc. Very likely the favored option since fundamental property of the SN mechanism as we know it (SASI, convection, KH, RT instabilities).

Velocity [km s-1]

L /

R

1D to multi-D: Hydro. Instabilities => Mixing

Shock fundamentally aspherical on large scale

Instabilities (RT, shocks, convection) => small scale clumps, fingers Main source of mixing in Type I b/c Sne

Strong RT when shock meets H/He interface in Type II SNe.

Impacts SN radiation for low-mass ejecta

Kifonidis et al. (2006)

Effects of Mixing: Numerical Exploration

Grid of models based on the SAME ejecta but with different levels of MIXING.

4.4M progenitor model -> Ejecta with 3M, 1.2B, 0.15M of 56Ni.

Increasing mixing from x0 to x4

Non-thermal processes

• Fundamental process associated with -ray and high-energy electrons

• Solution of the Spencer-Fano Equation (1954) following Kozma & Fransson (1992)

• Edep heat (thermal energy) & ionization/excitation (non-thermal) for all atom/ion levels

• Rates included in full non-LTE treatment (SEE equations).

• Key property: Non-thermal electrons deposit the bulk of their energy as heat ( with xe)

• Non-thermal excition/ionization: key for HeI in SNe Ib

Illustration with SN Ib model

Energydeposition

Excitation

IonizationHeat

Dashed: He

IonizationFraction

Impact on LC

LC: early peak, shortening/disparition of the post-breakout plateau.

Impact on colors

Mixing allows:

Early brightening + metal-line blanketing

Uniform heating, lower Tphot at peak, redder colors

Observations support such mixing

Impact on spectra

• HeI lines absent in unmixed low-mass models• HeI lines present in mixed low-mass models• Non-thermal + time-dependent effects infuence line strength & width!

Same Ekin but different line width!

Weak mixing Strong mixing

Effect of mixing in He-rich ejecta

Moderate MixingStrong Mixing

• No post-breakout plateau

• Prompt rise to peak

• Redder colors and broader spectral lines at peak

• Un-ambiguous presence of numerous HeI lines

• HeI/FeII/SiII in 6000A region but HeI stronger

• Numerous CI/OI/MgII/CaII lines at 7000-9000A

• 5.1M He-rich pre-SN star model

• 1.2B ejecta with 2 levels of mixing• Affects 56Ni strongly, He weakly • Non-thermal treatment

Strong Mixing: SN Ib

• post-breakout plateau.

• Re-brightening delayed

• Bluer colors and narrower spectral lines at peak

• HeI lines present but weak

• NaI/OI/HeI/FeII/SiII in 6000A region: Messy!

• Numerous CI/OI/MgII/CaII lines at 7000-9000A

Moderate Mixing: SN Ic

Models

Current status of SN simulations

• Code works for all SN types:

• CCSNe: II-P, II-pec, IIb, Ib, Ic (Dessart/Hillier 2010,2011ab,2012)

• SNe Ia (Dessart/Hillier/Blondin/Khokhlov)

• PISNe (Dessart/Hillier/Waldman/Livne)

Summary

• Original goals met. CMFGEN ‘‘works’’ for all SN types, fully non-LTE, fully time-dependent,

treatment of non-thermal processes, local/non-local energy deposition

• Non-LTE and steady-state good for testing, flexible. Suitable for modeling of young SNe II-P

• Accurate modeling requires time-dependent treatment (e.g. Dn/Dt) => Ionization freeze-out

• Non-thermal processes key in low-ionized conditions (SNe II, Ib, Ic).

• Higher physical consistency with “Full-Ejecta” simulations based on hydro inputs from pre-

SN evolution + explosion calculations

• Versatile tool for simultaneous spectroscopic and light-curve modeling of SNe.

• Support RSG for SNe II-P. MMS < 20M

• Future work: resolve Radius/luminosity issue. Effect of rotation and Z. Document nebular

spectra and [OI] for mass constraints.

• IIb/Ib/Ic LC support low-mass ejecta (20d rise, narrow peak, fast fading)

• Little H can produce a SN IIb – small difference with SNe Ib

• Non-thermal excitation of HeI supports mixing + low-mass ejecta in IIb and Ib

Summary (cont.)

• Mixing can influence Ib/Ic classification

• He deficiency not required for SN Ic type or LGRB/SNe.

• CCSNe all (?) come from low-mass massive stars (<20M):

1) Single stars produce RSG/BSG and II-P/II-pec

2) Binaries produce H-poor low-mass cores and IIb/Ib/Ic.

• What happens to higher mass 20-30M stars? Rotation may make them explode/die as WR

rather than RSG

• What happens to high-mass single WRs?

1) Fail to explode but why? Larger compactness? Stellar evolution wrong?

2) They explode but larger core Ebinding causes 56Ni fallback making a “dark” SN (detection?)

Spectro-polarimetry: Explore departures from spherical symmetry

• Inputs from non-LTE time-dependent SN II-P simulations (DH11a) • Fixed imposed asymmetry• Polarization due to electron scattering• Plateau-phase: Low-polarization due to strong cancellation effects, optical

depth effects, steep mass-density and/or electron-density distribution• For a fixed asymmetry, polarization jumps at onset of nebular phase• Nebular-phase: polarization commensurate to magnitude of asymmetry

Leonard et al. (2006) - Type IIP SN2004dj Dessart & Hillier (2011b) Oblate SN II-P ejecta seen edge-on

Treatment of Energy Deposition: -ray Monte-Carlo transport code

2-step decay chain: 56Ni 56Co 56Fe (1/2=6.1 and 77.27d)

+ decay, electron capture -rays, neutrinos, positrons, pairs

Monte-Carlo code to follow -rays trough ejecta subject to Compton Scattering and

photoelectric absorption

>>1: assume local deposition as heat. <1: non-local, non-thermal

Identification of CCSN progenitors on pre-explosion imagesM

attil

a et

al.

(200

8)

But most SNe have NO progenitor detection => Modeling needed

Modeling of Light Curves and Spectra: Some Features

• Radiative diffusion Time-dependent Transport, Energy/opacity problem

• Strong collisional proc. at depth LTE, blackbody SED

• Strong electron scattering non-LTE above Rphot, J ≠ B, flux ‘’dilution’’

• Large radii photospheric density is low (weak collisions) non-LTE effects

• line>> cont & Expansion P-Cygni profiles

• Large # of lines and large ejecta velocity Line overlap and blanketing

• Steep and/or Ne distribution Line formation, low linear polarization

• High V, low trec ~ R/V time-dependent effects (UC05, DH08)

• -rays High-energy e- Non-thermal effects (Lucy 1991)

Illustration with [OI] from CMFGEN simulations

• Broader/stronger [OI] line in more massive progenitors

• Sensitivity to gas temperature (collisional excitation)

• Modulations: 56Ni (heating), clumping, line overlap

Comparison with SN IIb 2011dh

SN2011dh; Arcavi et al (2011)

Dessart et al. (2011ab)

Encouraging match to SN 2011dh

with SN IIb model 4.4Msun model

of Dessart et al. (2011ab) for optical

and near-IR ranges.

Comparison with SN IIb 2011dh

SN2011dh; Marion et al. (ATel)

Encouraging match to SN 2011dh

with SN IIb model 4.4Msun model

of Dessart et al. (2011ab) for optical

and near-IR ranges.

Dessart et al. (2011ab)

Comparison to SN 2008D

SN 2008D: Blue featureless SED (intriguing...),

Appearence/persistence of HeI lines, Fast transition to nebular (red SED)

Structured SEDAppearance/persistance of HeI lines

(mixing, non-thermal effects),Slower transition to nebular (M, 56Ni))

SED too blue at late time (56Ni)

Strongly mixed SN Ib modelModeling with non-thermal processes.

• Non-thermal processes stronger in low-ionization conditions => stronger above the photosphere.

• Most energy is deposited thermally at depth though and affects the LC.

Results

Illustration of the gamma-ray volume of influence.