Edo Berger − Harvard University

Toward the Progenitors of Short-Duration Gamma-Ray Bursts

The Prompt Activity of Gamma-Ray Bursts: their Progenitors, Engines, and Radiation Mechanisms − March 2011

Outline

Kouveliotou et al. 1993

• The discovery of short GRB afterglows & host galaxies

• Redshift distribution & galaxy-scale environments

• Offset distribution & sub-galactic environments

• Evidence for Kicks?• Rates• Prompt emission & afterglows• mini-, macro-, kilo-novae?• Gravitational wavesBerger, E. 2011, New Astronomy Reviews, 51, 1

Short vs. Long GRBs

BATSE − short:long = 1:3

Swift − short:long = 1:10

1. Exclusively in star-forming galaxies

Long GRBs: The Death of Massive Stars

Fruchter et al. 2006; Wainwright, Berger, & Penprase 2007

2. Offsets trace star formation in an exponential disk

Fruchter et al. 2006

Long GRBs: The Death of Massive Stars

Bloom et al. 2002

3. Coincident with the brightest UV regions of

their hosts

Short GRB Progenitor ModelsNS-NS / NS-BH

•Broad delay-time distribution•Diverse environments / redshifts•“Kicks”•“Mini-SN”?•Gravitational wavesMagnetars

•Young systems•Star forming galaxies / nearby?•No “kicks”

Short GRB Progenitor ModelsWD/NS AIC

WD-WD merger

•Delayed magnetar / BH formation•Diverse environments •No “kicks”

Early “Smoking Guns”?

Castro-Tirado et al. 2005; Gehrels et al. 2005; Hjorth et al. 2005; Bloom et al. 2006; Prochaska et al. 2006

Berger et al. 2005

z = 0.257

Unambiguous association with an elliptical galaxy & no accompanying

supernova

⇔

z = 0.226

050509B

Afterglows Galore...

Soderberg et al. 2006

Berger et al. 2007

D’Avanzo et al. 2009

Berger et al. 2009

Fong et al. 2011

... and Host Galaxies Berger et al. 2007;

Berger 2009

Host Demographics

SF

Ell

?

“Host-less”

~1/2 of all short GRBs are located at z > 0.7 ⇒ 〈 age 〈≤ 7 Gyr

Berger et al. 2007; Berger 2009; Fong et al. 2011

z = 0.438

z = 0.922z = 1.130

z = 0.410

z = 0.827

Host Redshifts

z = 0.915

Progenitor Ages:

• P(τ) ∝ τn with n ~ −1*• τ ~ 3-4 Gyr, σ ~ 1 * MW: NS-NS systems have n ~

−1

Nakar 2006; Guetta et al. 2006; Berger et al. 2007

Host Redshifts

Host Galaxies: Star Formation Rates

Short GRB hosts have lower specific star formation rates than long GRB hosts; they

trace the general galaxy population

Berger 2009

Host Galaxies: Metallicities

Short GRB hosts have higher metallicities than long GRB hosts; they trace the general

galaxy population

Berger 2009

Host Galaxies: Stellar Masses

Leibler & Berger 2010

Short GRB hosts have higher stellar masses than long GRB hosts

Do Short GRB progenitors track stellar mass alone?

Stellar Population AgesLeibler & Berger 2010

Short GRB hosts (including star-forming) have older ages than long GRB hosts

τshort,SF ~ 0.3 Gyr

τshort,E ~ 3 Gyr

τlong ~ 60 Myr

Do Short GRBs Track Stellar Mass? SF?

Early-type hosts track stellar mass, but star-forming hosts have lower masses than expected;

star-forming dominate (1:1 expected)

Rell ~ 6×10−12 per M☉

Rsp ~ 2×10−11 per M☉

Leibler & Berger 2010

Sub-galactic Environments

Fong, Berger, & Fox 2010

•Are short GRBs associated with young or old stellar populations within their hosts?

•Is the distribution of offsets indicative of “kicks”?

Sub-galactic Environments: Light Fraction

Short GRBs trace low luminosity regions of their hosts; track optical (mass) better than

UV (SFR)

Fong, Berger, & Fox 2010Fruchter et al. 2006; Kelley et al. 2008

Sub-galactic Environments: Offsets

•Short GRB offsets are ~5x larger than for long GRBs•Good agreement with model predictions for NS-NS binaries

Fong, Berger, & Fox 2010

Is there Evidence for Large Kicks?

Berger 2010

Of 20 short GRBs with optical afterglows, 5 have no

coincident hosts to >26 mag.

• Underlying hosts >26 mag + fainter afterglows (high redshift)• No hosts + fainter afterglows (kicks / low density)

Is there Evidence for Large Kicks?

Berger 2010

high-z: same galaxies ⇒ bimodal redshift distribution

Offsets: low chance probability at ~10” ⇒ 50-100 kpc

z ~ 0.1-0.5

Is there Evidence for Large Kicks?

Berger 2010

⇒

Kicks?Berger 2010

Extension to larger offsets provides better agreement with the NS-NS merger models. Not expected in other models.

Rates

ℜSHB > 10 Gpc−3 yr−1

Nakar et al. 2007

Rell ~ 6×10−12 M☉−1

Rsp ~ 2×10−11 M☉−1

Leibler & Berger 2010

ρ* ~ 6×1017 M☉ Gpc−3

⇒

τ−1 > 0.4-1.2 Myr−1

ℜDNS ~ 17−290 Myr−1 (95%)Kalogera et al. 2004

Abdo et al. 2009b

Short GRB Prompt EmissionShort GRB090510 Long GRB09092B

Fermi prompt emission properties similar to long GRBs (GeV delay)

Abdo et al. 2009a

Nousek et al. 2006Barthelmy et al. 2005

Short GRB AfterglowsShort GRB050724 Long GRBs

X-ray afterglows are similar to those of long GRBs

Radio/optical afterglows similar to long GRBs

⇒ Afterglow & engine physics are similar to long GRBs

Berger et al. 2005

Short GRB Afterglows

Berger et al. 2000

Short GRB050724 Long GRB000301C

Berger et al. 2005

Short GRB Afterglows: 050724

Eγ,iso ≈ 4 × 1050 ergEK,iso ≈ 2 × 1051 ergθj > 25 degn ≈ 0.01−0.1 cm−3

εe ≈ εB ≈ 0.03

Barthelmy et al. 2005

•Extended X-ray emission•X-ray flare @ 1 day

Short GRB Afterglows: 051221

θj ≈ 7 degEγ ≈ 1.5 × 1049 ergEK ≈ 0.8 × 1049 ergn ≈ 1.5 × 10−3 cm−3

εe ≈ εB ≈ 0.2

Soderberg et al. 2006; Burrows et al. 2006

•X-ray plateau @ 0.1 days•Radio reverse shock•No supernova

Energy scale of short GRBs is generally lower than for long GRBs

Nakar 2007; Berger 2007; Gehrels et al. 2008; Nysewander et al.

2009

Short GRB Afterglows: Energetics

Short GRB Afterglows: CSM Density

Soderberg et al. 2006

The circumburst densities (~0.1 pc) are lower than for long GRBs

Mini-SN / Macronova / Kilonova?

neutron-rich ejecta:

•tidal tails (Mej < 0.1 M⊙)•AD outflows (Mej ~ 10-

3−10-2 M⊙)

Li & Paczynski 1998; Janka et al. 1999; Lee & Kluzniak 1999; Ruffert & Janka 2001; Rosswog et al. 2004; McLaughlin & Surman 2005; Rosswog 2005; Shibata & Taniguchi 2006; Metzger et al. 2008; Giacomazzo et al. 2009; Lee et al. 2009; Rezzolla et al. 2010

Decompressing neutron-rich ejecta ⇒ R-process (A ~ 100)vej ~ 0.1 cMej ~ 10−3 − 10−1 M⊙

fr.a. ~ 10−6

Lp ~ 1041 erg/stp ~ 1 day}

Courtesy: Brian Metzger

Mini-SN / Macronova / Kilonova?

Nakar & Piran 2011

Mildly-relativistic ejecta (or off-axis jet) will interact with ISM to produce a isotropic radio signal (cf radio SNe)

* at d ~ 300 Mpc

Strongly suppressed if n < 1 and/or β < 1 and/or E <1051

EVLA

Gravitational Waves

LIGO:Blind DNS detections: ~20 Mpc

Advanced LIGO: Blind DNS detections: ~0.2-0.3 GpcTriggered Short GRB: ~0.5-1.3 Gpc Nakar et al. 2006; Berger

et al. 2007

⇒ Detection rate of ~few per year

⇒ ⇒

⇒

Summary I

•The progenitors of short GRBs are not massive stars; they belong to an evolved population with a wide range of ages

•The short GRB rate in star-forming galaxies is elevated relative to that in ellipticals ⇒ A channel that tracks SF?

•If related to the star formation activity, the typical delay time is ~0.3 Gyr; the typical delay time in ellipticals is ~3 Gyr

•The short GRB volumetric rate, combined with the inferred rate per unit stellar mass, indicate a temporal rate of >1 Myr−1 per galaxy (consistent with the MW DNS rate)

Summary II

•The local environments of short GRBs exhibit:-large offsets (in agreement with NS-NS models)

-better correlation with optical (mass) than UV (SF)

-low density (~ 0.001-1 cm−3)

•Short GRBs with optical afterglows and no coincident hosts are likely due to kicks (bimodal redshift distribution?)

•Short and long GRB afterglows are similar, but lower E, n

•Orphan afterglow and mini-SN (optical/radio) detections extremely challenging ⇒ γ-ray triggers are essential!

•It is crucial to have a γ/X-ray mission capable of ~arcmin positions in conjunction with GW detectors.