fireworks from magnetar birth - purdue universitylyutikov/workshop16/talks/... · 2016-05-28 ·...

Fireworks from Magnetar Birth

In Collaboration with

GRB Magnetar Thinkshop Bormio, Italy - January 21, 2014

Indrek Vurm, Romain Hascoet, Andrei Beloborodov (Columbia), Tony Piro (Caltech) Eliot Quataert, Geoff Bower, Jon Arons (UC Berkeley)

Niccolo Bucciantini (INAF), Todd Thompson (OSU), Dimitrios Giannios (Purdue) Paul O’Brien, A. Levan (Leicester), A. Rowlinson (Amsterdam)

Brian Metzger Columbia University

Gamma-Ray Bursts (Long and Short) Super-Luminous

Supernovae Ω

µ

The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.

‘Millisecond Magnetars’


Ro ~ 1 for P ~ 1 ms


rotational energy:

Dessart et al. 2006

All neutron stars form as hot, differentially-rotating ‘proto-neutron stars’

Field amplification:

€

Log(Ro ≡ Pτc)

ΔΩ


Pizzolato et al. 2003

Rossby Number

•  Shear instabilities (talk by Zrake)

•  Magneto-rotational instability

•  α-Ω dynamo (Thompson & Duncan 1993)

Vc

Mag

netic

act

ivity

of l

ate

type

sta

rs

Signatures of Magnetar Birth spin-down luminosity

spin-down time

Bdi

p (G

)

Birth Period P0 (ms)

Galactic Magnetars

(Vin

k &

Kui

per 0

6)

Ero

t < 1

051 e

rg

: Ω

µ

€

Lsd =µ2Ω4

c 3 ≈ 6 ×1049 P1 ms⎛

⎝ ⎜

⎞

⎠ ⎟ −4 Bdip

1015 G⎛

⎝ ⎜

⎞

⎠ ⎟

2

erg s-1

€

τsd =E rot

Lsd

≈10 P0

1 ms⎛

⎝ ⎜

⎞

⎠ ⎟

2 Bdip

1015 G⎛

⎝ ⎜

⎞

⎠ ⎟

-2

min :

Signatures of Magnetar Birth

€

Lsd =µ2Ω4

c 3 ≈ 6 ×1049 P1 ms⎛

⎝ ⎜

⎞

⎠ ⎟ −4 Bdip

1015 G⎛

⎝ ⎜

⎞

⎠ ⎟

2

erg s-1

€

τsd =E rot

Lsd

≈10 P0

1 ms⎛

⎝ ⎜

⎞

⎠ ⎟

2 Bdip

1015 G⎛

⎝ ⎜

⎞

⎠ ⎟

-2

min

spin-down luminosity

spin-down time

€


Galactic Magnetars

(Vin

k &

Kui

per 0

6)

Gamma-Ray Burst !  Jet punches successfully through star

!  Lsd ~ Lγ ~ 1049-51 erg s-1

!  τsd ~ minutes-hours

Ero

t < 1

051 e

rg

:

:

Bdi

p (G

)


spin-down time

€


Galactic Magnetars

(Vin

k &

Kui

per 0

6)

SuperLuminous SNe


!  Lsd ~ Lγ ~ 1049-51 erg s-1


Super-Luminous SN !  Jet stifled, but optical SN powered diffusively

!  Lsd ~ LSN ~ 1043-45 erg s-1

!  τsd ~ week - months E

rot <

105

1 erg

:

€

Lsd =µ2Ω4

c 3 ≈ 6 ×1049 P1 ms⎛

⎝ ⎜

⎞

⎠ ⎟ −4 Bdip

1015 G⎛

⎝ ⎜

⎞

⎠ ⎟

2

erg s-1

€

τsd =E rot

Lsd

≈10 P0

1 ms⎛

⎝ ⎜

⎞

⎠ ⎟

2 Bdip

1015 G⎛

⎝ ⎜

⎞

⎠ ⎟

-2

min :

Bdi

p (G

)

•  Energies - Eγ ~ 1049-52 ergs

•  Rapid ~ms variability -

•  Duration - Tγ ~10-100 s

•  Steep decay phase

•  Narrowly collimated jet

•  Bulk Lorentz factor Γ ~ 100-1000 (Mjet < 10-5 M!)

•  Late activity (plateau & flaring)

Constraints on the GRB Central Engine

BH NS versus

Canonical GRB Lightcurve

from N

akar 07

SN 1998bw

GRB SNe are actually quite

successful!

Bright ⇒ MNi56 > 0.1 M!

Energetic ⇒ EKE ~ 1052 ergs

…but massive stars (ZAMS >25 M!) become black holes, right?

Des

sart

et a

l. 20

10

Binding Energy of Stellar Envelopes

normal SN

Des

sart

et a

l. 20

10

1052 ergs GRB SN

Binding Energy of Stellar Envelopes

normal SN

Core Collapse with Magnetic Fields & Rotation (e.g. LeBlanc & Wilson 1970)

Neu

tron

Sta

r Mas

s

€

˙ M IN€

˙ M OUT

Time

“Failed Collapsar”

Harder to Blow Up ⇒

“Fr

ee E

nerg

y” in

Diff

eren

tial R

otat

ion ⇒

⇐ Easier to Blow Up

Collapsar Progenitors?

Rot

atio

n P

erio

d at

Birt

h ⇒

Alternative View of the Fates of Massive Rotating Stars (Metzger et al. 2011; see also Dessart, O’Connor, & Ott 2012)

Neutrinos heat proto-NS atmosphere (e.g. νe + n ⇒ p + e-) ⇒ drives wind behind outgoing supernova shock (e.g. Qian & Woosley 96)

Neutrino-Heated Wind

Before SN Shock Launch After Shock Launch

Neutrino Driven Wind

Burrows, Hayes, & Fryxell 1995

€

˙ M ~ 10−4 Lν1052erg s-1

⎛

⎝ ⎜

⎞

⎠ ⎟

5 / 3εν

10 MeV⎛

⎝ ⎜

⎞

⎠ ⎟

10 / 3

M! s−1 ⇒ crucial to baryon loading

Effects of Strong Magnetic Fields “Helmet - Streamer”

Ω

•  Microphysics (EOS, ν Heating & Cooling) –  Important for B ≥ 1016 G (Duan & Qian 2005)

Effects of Strong Magnetic Fields “Helmet - Streamer”

€

B2

8π > 12ρ vr

2

Outflow Co-Rotates with Neutron Star when

RA Rheat

Ω

Top View

•  Microphysics (EOS, ν Heating & Cooling) –  Important for B ≥ 1016 G (Duan & Qian 2005)

•  Magneto-Centrifugal Slinging (Weber & Davis 1967; Thompson, Chang & Quataert 2004)

⇒

Magneto-Centrifugal Acceleration (“Beads

on a Wire”)

Regimes of Magnetized PNS Winds (B = 3×1014 G) N

eutr

ino

Lum

inos

ity (1

051 e

rg s

-1) Thermally-Driven

Magnetically-Driven, Ultra-Relativistic

Metzger, Thompson, Quataert 2007

Rotation Period (ms)

Regimes of Magnetized PNS Winds (B = 3×1014 G) N

eutr

ino

Lum

inos

ity (1

051 e

rg s

-1) Thermally-Driven

Magnetically-Driven, Ultra-Relativistic

Metzger, Thompson, Quataert 2007

Rotation Period (ms)

t ~ 100 s

t ~ 1 s

Evolution of Proto-Magnetar Outflows (BDM et al. 2007, 2011)

Initial rotation period P0 , dipole field Bdip & obliquity θdip

NS Cooling Luminosity

3D Magnetosphere Geometry (e.g. Bucciantini et al. 2006; Spitkovsky 2006)

Calculate:

€

Wind Power ˙ E (t), Mass Loss Rate ˙ M (t),

⇒ 'Magnetization' σ(t) ~ ˙ E ˙ M c 2 = Γmax (t)

In terms of

Roberts 2012

Neutrino Cooling Evolution

€

σ0 ~ Γmax =˙ E

˙ M c2 ∝B2Ω4

Lν5/3T10/3

€

spin - down power˙ E iso /1050 erg s-1€

magnetization σ0 ~ Γmax

Example Solution

increases as magnetar cools

RADIO X-RAYS OPTICAL

SNR

PWN

PULSAR

Multi-Wavelength Crab Nebula

3C58 (Chandra)

Collimation via Stellar Confinement

Collimation via Stellar Confinement

RADIO X-RAYS OPTICAL

SNR

PWN

PULSAR

Multi-Wavelength Crab Nebula

3C58 (Chandra)

Ω

Supernova remnant elongated by anisotropic magnetic stresses in pulsar nebula? (Begelman & Li 1992)

Outgoing SN shock VSN ~ 0.1 c

Fast Magnetar Wind Vw ~ c


Jet Formation via Stellar Confinement (Bucciantini et al. 2007, 08, 09; cf. Uzdensky & MacFadyen 07; Komissarov & Barkov 08)

Zoom Out

Jet power & mass-loading match (on average) outflow

from central magnetar 2D 3D

Porth, K

omissarov, &

Keppens 13

Kink Instability

Outflow becomes relativistic at t ~ 2 seconds; Jet breaks out of star at tbo ~ R!/βc ~ 10 seconds

Jet B

reak

-Out

Non

-Rel

ativ

istic

(σ0 <

1)

Rel

ativ

istic

(σ0 >

1)

€

σ0

€

˙ E iso /1050 erg s-1

Outflow becomes relativistic at t ~ 2 seconds; Jet breaks out of star at tbo ~ R!/βc ~ 10 seconds

Jet B

reak

-Out

Non

-Rel

ativ

istic

(σ0 <

1)

Rel

ativ

istic

(σ0 >

1)

€

σ0

€

˙ E iso /1050 erg s-1Je

t Bre

ak-O

ut

←GRB→

Central Engine GRB / Flaring

Relativistic Outflow (Γ >> 1)

Afterglow

1.  What is jet’s composition? (kinetic or magnetic?)

2.  Where is dissipation occurring? (photosphere? deceleration radius?)

3.  How is radiation generated? (synchrotron, IC, hadronic?)

~ 107 cm

Photospheric IC

GRB Emission - What, Where, How?

Metzger et al. 2011

€

˙ E jetJe

t Bre

ak-O

ut

Opt

ical

ly-T

hick

Optically-Thin

Time-Averaged Light Curve

Photospheric Dissipation (IC)

Metzger et al. 2011

€

˙ E jetJe

t Bre

ak-O

ut

Opt

ical

ly-T

hick

Optically-Thin

Time-Averaged Light Curve

Hot Electrons ⇒ IC Scattering (γ-rays)

and Synchrotron (optical)

E F E

(105

0 erg

s-1

)

Spectral Snapshots

t ~ 30 s

E (keV)

t ~ 15 s Synch

IC Tail BB

Epeak Evolution

!!!

Photospheric Dissipation (IC)

σavg-Lγ Correlation Av

e. M

agne

tizat

ion σ

avg

Ave. Wind Power (erg s-1)

σavg ∝ Lγ1-1.5

Prediction: More Luminous GRBs

⇔ Higher Γ

Epeak∝ Liso0.5

Peak Liso (erg s-1)

Aver

age

E pea

k (k

eV)

Epeak∝ Liso0.11 σ0

0.2 (Giannios 2012)

Consistent with Epeak ∝ Eiso

0.4 (Amati

+02)

Epeak ∝ Liso0.5

(Yonetoku+04)

And Γ∝ Eiso0.3

(Liang+10) Correlations

Assuming Magnetic Dissipation Model

End of the GRB = Neutrino Transparency?

Ultra High-σ Outflows ⇒ - Acceleration is Inefficient (e.g. Tchekhovskoy et al. 2009) - Internal Shocks are Weak (e.g. Kennel & Coroniti 1984) - Reconnection is Slow (e.g. Drenkahn & Spruit 2002)

TGRB ~ Tν thin ~ 20 - 100 s

€

σ0

€

˙ E iso /1050 erg s-1

←GRB→

baryons e-/e+ pairs

Steep Decline

End of the GRB = Neutrino Transparency?

Ultra High-σ Outflows ⇒ - Acceleration is Inefficient (e.g. Tchekhovskoy et al. 2009) - Internal Shocks are Weak (e.g. Kennel & Coroniti 1984) - Reconnection is Slow (e.g. Drenkahn & Spruit 2002)

TGRB ~ Tν thin ~ 20 - 100 s

€

σ0

€

˙ E iso /1050 erg s-1

←GRB→

baryons e-/e+ pairs

Low plateau efficiency consistent with Lu & Zhang 2014

←GRB→

τSD

e.g. Zhang & Meszaros 2001; Troja et al. 2007; Yu et al. 2009; Lyons et al. 2010

Late-Time (Force-Free)

Spin-Down

€

˙ E iso /1050 erg s-1

←GRB→

τSD

Willi

ngal

e et

al.

2007

`Plateau’

Time after trigger (s)

X-ray Afterglow

e.g. Zhang & Meszaros 2001; Troja et al. 2007; Yu et al. 2009; Lyons et al. 2010; Rowlinson et al. 2010, 2013; Gompertz et al. 2013

Late-Time (Force-Free)

Spin-Down

€

˙ E iso /1050 erg s-1

Data from Lyons et al. 2010

Plateau Duration - Luminosity Correlation

`Plateau’ Luminosity

Spi

n-D

own

Tim

esca

le

A Diversity of Magnetar Birth

P0 (ms)

Bdi

p (G

) Classical GRB Eγ~1050-52 ergs,

τjet < 1, Γ ~ 102-103


P0 (ms)

Classical GRB Eγ~1050-52 ergs,

τjet < 1, Γ ~ 102-103

Thermal-Rich GRB (XRF?) Eγ~1050 ergs, τjet ~ 1, Γ < 10

Bdi

p (G

)


Buried Jet

P0 (ms)

Classical GRB Eγ~1050-52 ergs,

τjet < 1, Γ ~ 102-103

Very Luminous SNe? (Kasen & Bildsten 10; Woosley 10)

Thermal-Rich GRB (XRF?) Eγ~1050 ergs, τjet ~ 1, Γ < 10

Bdi

p (G

)

Galactic Magnetars?

Observational Tests & Constraints •  Max Energy* - EKE+Eγ < 3×1052 ergs *subject to uncertainties in afterglow modeling. (e.g. Zhang & MacFadyen 09).

•  Long GRB always accompanied by bright, energetic

- Consistent with observations thus far (Woosley & Bloom 2006).

•  Γ increases during GRB and correlates with Eγ - translate jet luminosity/magnetization into unique prediction for gamma-ray light curves and spectra.

Cenko et al. 2011


spin-down time

€


Galactic Magnetars

(Vin

k &

Kui

per 0

6)

SuperLuminous SNe


!  Lsd ~ Lγ ~ 1049-51 erg s-1


Super-Luminous SN !  Jet stifled, but optical SN powered diffusively

!  Lsd ~ LSN ~ 1043-45 erg s-1

!  τsd ~ week - months E

rot <

105

1 erg

:

€

Lsd =µ2Ω4

c 3 ≈ 6 ×1049 P1 ms⎛

⎝ ⎜

⎞

⎠ ⎟ −4 Bdip

1015 G⎛

⎝ ⎜

⎞

⎠ ⎟

2

erg s-1

€

τsd =E rot

Lsd

≈10 P0

1 ms⎛

⎝ ⎜

⎞

⎠ ⎟

2 Bdip

1015 G⎛

⎝ ⎜

⎞

⎠ ⎟

-2

min :

Bdi

p (G

)

"  GRB Duration ~ 10 - 100 seconds & Steep Decay Phase - Time for NS to become transparent to neutrinos (end of ν-wind)

"  GRB Energies EGRB ~ 1050-52 ergs

- Rotational energy lost in ~10-100 s

"  Ultra-Relativistic Outflow with Γ ~ 100-1000 - Mass loading set by physics of neutrino heating (not fine-tuned).

"  Jet Collimation - Star confines and redirects magnetar outflow into jet "  Association with Energetic Core Collapse Supernovae - Erot~ESN~1052 ergs - MHD-powered SN associated w magnetar birth.

"  Late-Time Central Engine Activity - Residual rotational (plateau) or magnetic energy (flares)?

Summary of the Proto-Magnetar Model for GRBs

Hydrogen-Poor ‘SuperLuminous’ Supernovae Quimby+07, Barbary+09, Pastorello+10, Chomiuk+11, Leloudas+12, Berger+12, Lunnan+13, Inserra+13; Nicholl+13; McCrum+14

•  Lpeak > 1044 erg s-1, Erad~1050-51 ergs (10-100 × normal SNe) •  UV-rich spectrum with intermediate mass elements

•  Faint metal-poor host galaxies, similar to long GRBs (Quimby+11, Neill+11, Chomiuk+11, Chen+13; but see Berger+13, Chornock+13)

•  Competing models: circumstellar interaction vs. central engine

Qui

mby

et a

l. 20

11

Millisecond Magnetar-Powered Supernovae (Kasen & Bildsten 2010; Woosley 2010; Dessart et al. 2011)

•  Can reproduce diversity of rise times and peak luminosities (hard to test)

•  Difficult to distinguish from other ‘hidden’ energy sources (optically-thick CSM interaction)

•  Assumes pulsar luminosity thermalized Reality: Poynting flux ⇒ e+/- ⇒ non-thermal radiation ⇒ thermal radiation

Kas

en &

Bild

sten

2010

PROS •  SN luminosity increased if pulsar spin-down time ~ optical peak tpeak ⇒ Bdip ~ 1013-14 G

•  Explains similar host galaxies to long GRBs (both require rapidly rotating progenitor)

•  Can reproduce diversity of rise times and peak luminosities

CONS

How Supernovae Shine (Arnett 1982) SN ejecta w mass M, velocity, & opacity κ

M V

€

ρ =M

4π3R

3

} R

€

τ ~ κρR

€

t diff ~ τ Rc

€

R = v t

Light escapes when t > tdiff ⇒

€

t > month v

104 km s-1

⎛

⎝ ⎜

⎞

⎠ ⎟ −1/ 2 M

3M!

⎛

⎝ ⎜

⎞

⎠ ⎟

1/ 2κ

0.1 cm2g−1

⎛

⎝ ⎜

⎞

⎠ ⎟

1/ 2

Optical (κ ~ 0.1 cm2 g-1) ⇒ t ~ month

Hard X-ray ~ 10 keV (κ ~ 30 cm2 g-1) ⇒ t ~ few years

Soft X-ray ~0.1 keV (κ ~ 105 cm2 g-1 ) ⇒ t ~ 100 years

UV/X-ray opacity of neutral oxygen

κ

κopt

X-ray Ionization Break-Out

1.  Pulsar inflates cavity (pulsar wind nebula)

2.  Nebula X-rays ionize inner exposed surface of ejecta

3.  Ionization front reaches outer surface - X-rays escape to observer.

neutral SN ejecta

nebula (e+/- pairs, photons)

ionization front(s)

(BDM, Vurm, Hascoet & Beloborodov 2013)

Lum

inos

ity (e

rg s-1

) Optical

X-ray

Time (days)

Evolution of Millisecond Pulsar Wind Nebulae

Non-Thermal UV / X-rays Source: cooling e+/- pairs (pulsar)

Sinks: PdV work, absorption by ejecta walls Thermal Bath (Optical) Source: re-emission of X-rays by ejecta walls

Sinks: PdV work, radiative diffusion

Ejecta Ionization State - Balance photo-ionization with recombination in

ionized layer(s)

-  Sets ejecta albedo (thermalization efficiency)

(BDM, Vurm, Hascoet & Beloborodov 2013)

analogy to AGN accretion disks

Example: B = 1013 G, P = 1 ms, Mej = 3 M! (B

DM

, Vur

m, H

asco

et &

Bel

obor

odov

201

3)

Ioni

zatio

n D

epth

Break-Out

Fe20+

O1+ O2+ O3+

Example: B = 1013 G, P = 1 ms, Mej = 3 M! (B

DM

, Vur

m, H

asco

et &

Bel

obor

odov

201

3)

Ioni

zatio

n D

epth

Break-Out

Fe20+

O1+ O2+ O3+

Initial Spin Period P (ms) Initial Spin Period P (ms)

Mag

netic

Fie

ld B

d (G

)

Mag

netic

Fie

ld B

d (G

)

Superluminous X-rays from a Superluminous SN C

hatz

opou

los e

t al.

2013

(cf.

Barb

ary

et a

l. 20

09) (Levan, Read, BDM, Wheatly, Tanvir 2013)

MOS1 MOS2 pn

Lx ~ 1045 ergs s-1

X-rays!

0.2-2 keV

Superluminous X-rays from a Superluminous SN C

hatz

opou

los e

t al.

2013

(cf.

Barb

ary

et a

l. 20

09)

!  No detections from other SLSNe

!  Upper limits Lx < 1042-1044 erg s-1 on timescales < 70 days (usually too early!)

!  Future: X-ray follow-up after optical peak confirm or constrain pulsar model for SLSNe

(Levan, Read, BDM, Wheatly, Tanvir 2013)

MOS1 MOS2 pn

Lx ~ 1045 ergs s-1

Leva

n et

al.

2013

X-rays!

SLSN-I

0.2-2 keV

Summary •  Rapid (millisecond) birth period may be key to generating large

scale magnetar-strength B fields. •  Powerful outflow (τsd ~ min-hour) ⇒ relativistic jet ⇒ GRB

–  Baryon loading set by neutrino heating above magnetar surface. –  Accounts for GRB energetics, Lorentz factors, duration, collimation, late

activity; natural association with energetic supernovae. –  Key issues: stability of 3D jet and predicted rise in magnetization during

GRB. •  Weaker outflow (τsd~ weeks) ⇒ jet trapped ⇒ SuperLuminous SN

–  Previous models assume pulsar wind thermalizes with 100% efficiency. –  We have developed a model for the evolution of young msPWNe that couples

X-ray and thermal radiation via interaction with ejecta walls –  Pulsar wind ⇒ e+/- pairs ⇒ X-rays ⇒ thermal (optical) photons ⇒ observer

(optical SN) –  Nebular UV/X-rays can re-ionize ejecta within months of optical peak

(‘Ionization Break-Out’), allowing escape of high energy radiation.

fireworks from magnetar birth - purdue universitylyutikov/workshop16/talks/... · 2016-05-28 ·...

Documents