33
Agnieszka Janiuk N. Copernicus Astronomical Center, Warsaw Gamma Ray Bursts from Collapsing Massive Stars Collaborations: Collaborations: R. Moderski (CAMK), D. Proga(UNLV), Y. Yuan R. Moderski (CAMK), D. Proga(UNLV), Y. Yuan (ChAS), R. Perna (JILA), T. Di Matteo (ChAS), R. Perna (JILA), T. Di Matteo (CMU), B. Czerny (CAMK), D. Cline (UCLA), (CMU), B. Czerny (CAMK), D. Cline (UCLA), S. Otwinowski (CERN), C. Matthey (CERN) S. Otwinowski (CERN), C. Matthey (CERN)

Agnieszka Janiuk

  • Upload
    lovey

  • View
    33

  • Download
    0

Embed Size (px)

DESCRIPTION

Agnieszka Janiuk. N. Copernicus Astronomical Center, Warsaw. Gamma Ray Bursts from Collapsing Massive Stars. Collaborations: R. Moderski (CAMK), D. Proga(UNLV), Y. Yuan (ChAS), R. Perna (JILA), T. Di Matteo (CMU), B. Czerny (CAMK), D. Cline (UCLA), S. Otwinowski (CERN), C. Matthey (CERN). - PowerPoint PPT Presentation

Citation preview

Page 1: Agnieszka Janiuk

Agnieszka JaniukN. Copernicus Astronomical Center, Warsaw

Gamma Ray Bursts from Collapsing Massive StarsCollaborations:Collaborations:R. Moderski (CAMK), D. Proga(UNLV), Y. Yuan (ChAS), R. Moderski (CAMK), D. Proga(UNLV), Y. Yuan (ChAS), R. Perna (JILA), T. Di Matteo (CMU), B. Czerny R. Perna (JILA), T. Di Matteo (CMU), B. Czerny (CAMK), D. Cline (UCLA), S. Otwinowski (CERN), C. (CAMK), D. Cline (UCLA), S. Otwinowski (CERN), C. Matthey (CERN)Matthey (CERN)

Page 2: Agnieszka Janiuk

1014 cm

Gamma rays are believed to come from the internal shocks, produced in the relativistic (Γ>100) fireball.

To emit fireball, the engine must be To emit fireball, the engine must be very energetic. To produce shocks, very energetic. To produce shocks,

the engine must be active and the engine must be active and variable for a long timevariable for a long time

Page 3: Agnieszka Janiuk

Digression: what makes the jet?• Jets are ubiquitous in nature: AGNs,

QSOs, XRBs, YSOs, GRBs...• They are not required by any physical

law (such as energy conservation).• The 3 proposed mechanisms of jet

acceleration:– Radiation pressure– Thermal expansion– Magnetic fields and rotation

• Jet is domineted by– Poynting flux (small scales)– Matter (large scales)

• Jets are collimated by:– Accretion disk/coronal walls– Pressure gradients in the

environment– Surrounding matter-dominated jet– Poynting-jet is able to collimate itself

through the toroidal B fieldFragile, 2008 (arXiv:0810.0526)

Page 4: Agnieszka Janiuk

The model of a central engine for GRB must answer, which astrophysical process produces the relativistic fireball that emits gamma-rays.

Important constraintsImportant constraints (Piran 2005) (Piran 2005)::

Timescales and variability: dt/T ~ 10-3 – 10-4 for 80% of bursts Short and long GRB dichotomyShort - hard GRBs (T

90<2 s)

Long -soft GRBs (T90

> 2s) Energy (significant fraction of the binding energy for compact object) Collimation (1o < θ < 20o) Rates (about 3 ×10-5 per year per galaxy)

Page 5: Agnieszka Janiuk

To emit fireball, the engine To emit fireball, the engine must be very energetic. To must be very energetic. To produce shocks, the engine produce shocks, the engine must be active and variable must be active and variable for a long timefor a long time.

The most popular model invokes the internal shocks

in the jet that produce gamma rays and variability (Sari & Piran 1997).

Also, variability can be well reproduced with a shot-gun model (Heinz & Begelman 1999). Janiuk, Czerny, Moderski et al. (2006)

Kinematic jet model: theoretical lightcurves of long Kinematic jet model: theoretical lightcurves of long GRBs, depending on the observer’s viewing angleGRBs, depending on the observer’s viewing angle

Tested against observations: lightcurves, PDS Tested against observations: lightcurves, PDS spectra; Prokopiuk & Janiuk, in prep.spectra; Prokopiuk & Janiuk, in prep.

Page 6: Agnieszka Janiuk

Collapse of massive star favored for long GRBs:- associacion with star forming galaxies (e.g. Fruchter et al. 2006)- concurrent “SN-like” outbursts (Bloom et al. 1999; Stanek et al. 2003)- redshift distribution follows the star formation rate (Coward 2007)

Page 7: Agnieszka Janiuk

type I: rapid lightcurve evolution- - type Ia: standard; - - type Ib: He lines produced in the massive ejecta, by non-thermal excitation by fast particles emitted by the (56)Ni -> (56)Co-> (56)Fe decay.

- type Ic: progenitor must be either an extreme WR star, or a binary (Nomoto 1995)- type II: progenitor is a massive red giant

Supernovae

Page 8: Agnieszka Janiuk

Supernovae observed in Supernovae observed in associacion with GRBsassociacion with GRBs

- SN1998 bw: GRB 980425- SN 2003 dh: GRB 030329- SN 2003 lw: GRB 031203- SN 2006 aj: XRF 060218

All of these are Type IcAll have broad line spectra -> ejection velocities ~ 50,000 km/s

They account for 20% of the BL SN Ic = 2% of all SN Ic

Page 9: Agnieszka Janiuk

Hypernova: - very high expansion velocity- bright luminosity- postulated to be an energetic outburst produced by a collapsar (Woosley 1993; Paczyński 1998)- very strong explosion energy (> few x 1051 ergs)- strong evidence for assymetry (Nomoto et al. 2005)- massive star models fit well the observed hypernovae ( Mazzali et al. 2006)- large uncertainty in modeling due to the initial mass function of massive stars (5-40% core collapse SN form the black hole; Fryer & Kalogera 2001)

Eta Carinae: future candidate for hypernova

Page 10: Agnieszka Janiuk

Hypernovae are rare (about 1000 times less frequent than normal SN; (Soderberg et al. 2006)

All hypernovae have been classified as Ib/c SN (no H lines, nor He lines in the spectra); probably a subset of them

RatesRates of of Supernova vs. Hypernova

Rate of all core-collapse SN: 6x10-3 /yr/galaxy ( Fryer et al. 2007)

Type Ib/c are 15% of all core collapse

Hypernovae are 5-10% of observed type I b/c

1-10% of SN Ib/c can be associated with GRBs; this coincides with that of hypernovae

Rate of all core-collapse increases with redshift (no specific data for I b/c or hypernova)

Page 11: Agnieszka Janiuk

The collapsar

Woosley (1993): SN Type Ib 'failed' because of a fast rotation of the Wolf-Rayet star

Paczyński (1998): some GRBs must be linked to the cataclysmic deaths of massive stars -> hypernovae

MacFadyen & Woosley (1999) and follow up works: hydrodynamical computations of the relativistic jet propagation through the stellar envelope

GRB progenitors: the most massive stars, that fail to produce an explosion under the standard core-collapse supernova

Two reasosns for SN to fail (Fryer 1999): Large ram pressure at the top of the convective zone Large binding energy for the most massive stars

MacFadyen & Woosley (1999)

Page 12: Agnieszka Janiuk

Must form a black hole in the center of the star Must produce

sufficient angular momentum to form a disk around black hole Must eject the

hydrogen envelope, so that the jet can punch out of the star

The collapsar engine of a GRB

Page 13: Agnieszka Janiuk

Most massive stars (Mass > 20 solar masses; Hirshi et al. 2004)Wolf-Rayet stars: have lost the H envelope due to strong windsSingle stars: only fast rotating stars above solar metallicity produce strong shocks and eject lots of nickel (Heger at al. 2003)Fast rotating stars can mix their envelopes, burning effectively H into He (Yoon & Langer 2005)Binary stars: mass transfer can eject matter and lead to He star formation. Possibly, >75% of all massive stars are in close binaries (Kobulnicky et al. 2006)

SN 2008D

Progenitors of the I b/c Supernovae: single or binary stars?

Page 14: Agnieszka Janiuk

Single stars: only fast rotating stars above solar metallicity produce strong shocks and eject lots of nickel (Heger at al. 2003)

This GRB rate mustbe lower by a factorIndicating the fractionof stars that retain large angularmomentum

Page 15: Agnieszka Janiuk

Metallicity measurements

Wind mass loss sensitive to metallicity

At lower metallicities, weaker winds allow more massive cores => GRBs probably will not occur above solar metallicity

Metallicity measurements: Absorption lines in the GRB

afterglows Emission lines of HII regions in the

GRB host galaxy Interstellar extinction in the host

galaxy Morphology of the host galaxy, e.g.

Compared to SMC/LMC

Nebula NGC 2359

There is no consistent picture: direct measurements argue for higher Z, while indirect measurements indicate lower Z.

Page 16: Agnieszka Janiuk

Progenitors of Hypernovae Most of the currently discussed progenitors do not

distinguish between fallback and direct collapse black holes

GRBs probably will not occur at solar metallicity, if we need a direct collapse to black hole. At lower metallicities, weaker winds allow more massive cores.

Below ~0.4 ZSun, the stars cannot loose the He envelope (Heger 2003).

Star is either born rotating rapidly, or is spun up by interaction (tidal forces, merger). In binaries, the companion is used to strip off the hydrogen envelope without the angular momentum loss.

Single stars can also loose the H envelope because of mixing and burning to He (Yoon & Langer 2002). But if the He envelope is also lost, these models are ruled out.

„Constraints are more restrictive for single-star models, but without better understanding of winds we cannot say more” (Fryer et al. 2007).

WR124

Page 17: Agnieszka Janiuk

How long is a long GRB?

Chemical composition and density distribution in the pre-SN star (Woosley & Weaver 1995)

(Janiuk, Proga, Moderski. 2008a, 2008b)

Page 18: Agnieszka Janiuk

How the pre-collapse star rotates?

The distribution of specific angular momentum in the pre-SN star unknown.

Stellar evolution models: Neglect centrifugal forces Do not accurately treat the angular momentum transport through

magnetic fields Sensitive to the loss of ang. momentum through wind

Some assumptions we have made:

Polar angle dependence (differential rotation)

Radius dependence (rigid rotation, with a possible cut-off on lspec)

Constant ratio of centrifugal to gravitational forceslspec

= l0 (1-cos θ)

lspec

= l0 (r/r

in)sin2θ

(Janiuk et al. 2008a, 2008b)

Page 19: Agnieszka Janiuk

Conditions for torus existence

The rotation must prevent the envelope material from the radial infall onto BH.

BUT: – Black hole mass is growing fast (accretion rate of 0.01-1 M

sun/s)

– Spin can be changing

=> the GRB is emitted only until l>lcrit is satisfied.

Specific angular momentum l

spec > l

crit = 2GM

BH/c (2-A+2(1-A)1/2)1/2

Page 20: Agnieszka Janiuk

The black hole grows due to accretion

The time evolution of the collapsar => iterative procedure1. BH mass = iron core mass2. Envelope schells accrete3. Check for conditions given by the changing BH mass and

spinVarious possible accretion

scenarios

Page 21: Agnieszka Janiuk

GRB requires: large accretion rate and spinning BH

Schwarzschild and Kerr BH case:Janiuk & Proga, 2008, ApJ, 675, 519;Janiuk, Moderski & Proga, 2008, ApJ, in press;

Page 22: Agnieszka Janiuk

Hyperaccretion: neutrino-cooled disk

Electron-positron capture and beta-decay

p + e- → n + νen + e+ → p + νen → p + e- + νe

Thermal emissione+ + e- → νi + νin + n → n + n + νi + νiγ →νe + νe

-

-

-

-

-

(Popham, Woosley & Fryer 1999; Di Matteo, Perna & Narayan 2002; Kohri & Mineshige 2002; Janiuk et al. 2004; Kohri, Narayan & Piran 2005; Janiuk et al. 2007; Chen & Beloborodov 2007)

- Cooling mechanisms: neutrino emission,advection, Helium photodisintegration, radiation- Neutrinos can be absorbed and scattered- Equation of state should treat the species under the condition of reactions equilibrium and supplemented by the charge neyutrality condition

-

np e-

e+α

ν

Page 23: Agnieszka Janiuk
Page 24: Agnieszka Janiuk

The disk accreting at rates > 0.1 MSun s-1 is so hot and dense (T~1010-1011 K, ρ~1010-1012 g cm-3) that the plasma is totally opaque to photons, and neutrinos can also be trapped

Energy from the disk can be extracted by neutrino annihilation

Alternatively, energy can be extracted by the magnetic field and spinning black hole (the Blandford-Znajek mechanism)

Chen & Beloborodov (2007)

Page 25: Agnieszka Janiuk

Efficiency of neutrinos depends on initial accretion rate, and decreases in time

Neutrino annihilation inefficient when accretion rate < ~0.01 Msun

/s.This will slightly depend on viscosity and black hole spin.

Janiuk et al. (2004)

Chen & Beloborodov (2007)

Page 26: Agnieszka Janiuk

We end up with three kinds of jets from the collapsar:

0-1.5 s 0-130 s 0-430 s

Precursor jet,powered by v-vlarge m, small A

First jet, powered by both v-v and BH rotationlarge m, large A

Second jet, powered by BH rotationsmall m, large A

-

-

.

Page 27: Agnieszka Janiuk

GRB 0803319B

Brightest optical counterpart: m

v = 5.3

Empirical model of a 2-component jet fitted with two opening angles (Racusin et al. 2008)

Lazzati (2005) Precursors found in ~20% of BATSE sample

Image from “Pi of the sky”, http://grb.fuw.edu.pl

Page 28: Agnieszka Janiuk

Instabilities in the accreting torus: possible mechanism of causing a long time gap between the precursor and the burst, or the short-term variability seen in the prompt phase (Wang & Meszaros 2007).

Precursor phase in the prompt emission seen in some GRBs, might be produced by the jet breaking through envelope (Paczyński 1998; Ramirez-Ruiz et al. 2002)

Recent hydro Simulations by Morsony, Lazzati & Begelman (2007) found three distinct phases during the jet propagation: precurosr jet, shocked phase and unshocked phase.

Density, pressure and gamma_inf at time 30 s.

Cocoon: high ρ, high P, low Γ;

Precursor: high Γ, off-axis;

Shocked jet: low ρ, high P, high ρ;

Unshocked jet: low ρ, low P, high Γ

Page 29: Agnieszka Janiuk

Instabilities in the accretion disk:

- May be related to the late-time activity of the GRB (such as X-ray flares; e.g. Perna et al. 2006)

- proposed as the sources of gravitational waves, that may probe the angular momentum of the collapsing star (Fryer et al. 2002)

- Black hole spin can be coupled to the disk, enhancing the strength of the instability, then possibly detectable by LIGO (van Putten 2005)

Page 30: Agnieszka Janiuk

Thermal instability: the local density and pressure drops, while the temperature increases.

Our solution is based on the detailed treatment of the EOS, coupling the beta-equilibrium and the neutrino trapping effects, as well as including the information of the chemical composition in the process of Helium photodisintegration.

Janiuk et al. (2007)

Page 31: Agnieszka Janiuk

GRB long durations may provide constraints for the rotation law in the pre-SN star.

The minimum accretion rate limit for the neutrino-powered jets, in the Schwarzschild black hole models, results in GRB durations up to 40-100 s.

The minimum accretion rate and BH spin limit, for jets powered by both neutrinos and black hole spin, results in GRB durations up to 50-130 s.

The above values will be smaller if the H/He envelope was already stripped

In the Kerr black hole models, we find the solutions corresponding to three kinds of jets: precursor jet, early jet and late jet, powered by different mechanisms. Possibly, the opening angle of these jets is changing, which would have some observational consequences.

The instabilities in the accreting torus play important role for the observed emission

Summary

Page 32: Agnieszka Janiuk

Constraints on the GRB progenitor from observations and SN models

Type Ic SN => progenitor must loose the H and most of He envelope

Occur in the brightest parts of galaxies => come from the most massive stars

Occur in metallicities from 0.01 to 1 => single star models strongly constrained

Single star models may require mixing to burn H into He effectively

Binary star models fit better to the observational constraints

Page 33: Agnieszka Janiuk

Thank you