Radio Searches of GW CounterpartsCurrent and future capabilities

Dale A. FrailNational Radio Astronomy Observatory

Talk outline.

• What is the expected strength of the radio signal?– Afterglow component. Early and Late. (robust)

– Prompt counterpart (speculative).

• How do we detect the radio signal of a GW trigger?– The quiescent and transient radio sky. A primer.

– Current and future radio facilities.

– Three search strategies (in order of probability of success)

• What follow-up would we want to do?

• What can we be doing today to help the field?

2

3

νm≈Γ4 (i.e radio AG traces trans-relativistic ejecta)

Afterglow Radio Signal – Robust• Early radio emission (~days, weeks)

– SHB have lower average redshifts, energy and circumburst densities compared to long duration GRBs

– Only two SHB detected in radio out of ~25 Swift events.

GRB 050724 (z=0.257) and GRB 051221 (z=0.546)

– Best estimate is <Fradio>=100 μJy and <z>=0.5

– Predicts 10’s mJy at 1-10 GHz for d=200 Mpc

Afterglow Radio Signal – Robust

Van Eerten et al. (2010).

Early, on-axis

Early, on-axis

Late, off-axisLate, off-axis

L-GRB

Late-time radio detects AG independent of beaming

Afterglow Radio Signal – Robust• Early radio emission (~days, weeks)

– SHB have lower average redshifts, energy and circumburst densities compared to long duration GRBs

– Only two SHB detected in radio out of ~25 Swift events.

GRB 050724 (z=0.257) and GRB 051221 (z=0.546)

– Best estimate is <Fradio>=100 μJy and <z>=0.5

– Predicts 10’s mJy at 1-10 GHz for d=200 Mpc

• Late-time radio emission (~months)– Outflow expands, becomes quasi-isotropic and non-

relativistic. A late-time radio turn on independent of original jet direction.

– For reasonable SHB parameters t=30 days, F=0.3 mJy at 1.4 GHz at 300 Mpc (Nakar et al. in prep)

Prompt Radio Signal – Speculative

• Gravitationally excited MHD waves (Postnov & Pshirkov 2009)

– Predicts 12.5 Kilo-Jy at 100 MHz for d=200 Mpc

• Rotational energy of post-merger object (Moortgat & Kuijpers 2004)

– Predicts 50 Mega-Jy at 30 MHz for d=200 Mpc

• Emission from PSR-like magnetosphere (Hansen & Lyutikov 2001)– Predicts 1 milli-Jy at 400 MHz for d=200 Mpc

• “Back of the envelope” approach– Radio emission is seen in all high energy processes where there

are relativistic particles and magnetic fields– Assume that 10-6 of energy of a SHB goes into a prompt radio

signal– Average fluence for SHB is 10-6 erg cm-2. Duration 0.1 s– Predicts 1 kilo-Jy at 1 GHz

Quiescent and Transient Radio Sky. Primer.• Isotropic source distribution on sky

– Above 1 mJy source populations are AGN dominated – Below 1 mJy star-formation galaxies start to emerge

• The transient radio sky is quiet– GHz flux density range 0.1 mJy to 10 Jy is well studied

by several (heterogeneous surveys)– Transients are 10-3 to 10-4 of quiescent population

• e.g. Levinson et al. NVSS/FIRST comparison• Ofek et al. survey

• Important implication is that radio false EM-GW detection rate will be small (<0.1 deg2 at 1 mJy)

… and any background events are likely to be AGN, and hence easily filtered out.

The Quiescent Radio Sky is Isotropic

9

J. Condon

Quiescent and Transient Radio Sky• Isotropic source distribution on sky

– Above 1 mJy source populations are AGN dominated – Below 1 mJy star-formation galaxies start to emerge

• The transient radio sky is quiet– GHz flux density range 0.1 mJy to 10 Jy is well studied

by several (heterogeneous surveys)– Transients are 10-3 to 10-4 of quiescent population

• e.g. Levinson et al. NVSS/FIRST comparison• Ofek et al. survey

• Important implication is that radio false EM-GW detection rate will be small (<0.1 deg2 at 1 mJy)

… and any background events are likely to be AGN, and hence easily filtered out.

The Transient Radio Sky is Quiet

11

Ofek et al. (2011)

Quiescent and Transient Radio Sky• Isotropic source distribution on sky

– Above 1 mJy source populations are AGN dominated – Below 1 mJy star-formation galaxies start to emerge

• The transient radio sky is quiet– GHz flux density range 0.1 mJy to 10 Jy is well studied

by several (heterogeneous surveys)– Transients are 10-3 to 10-4 of quiescent population

• e.g. Levinson et al. NVSS/FIRST comparison• Ofek et al. survey

• Important implication is that radio false EM-GW detection rate will be small (<0.1 deg2 at 1 mJy)

… and any background events are likely to be AGN, and hence easily filtered out.

Radio facilities for GW-EM Counterpart Searches: 2011 and Beyond

13

EVLA

WSRT/Apertif

LOFAR

ASKAP

MWAMeerKAT

Radio facilities for GW-EM Counterpart Searches

RadioFacility

ObservingFreq.

Field of View

1 hr rms

Beam StartDate

ASKAP 1.4 GHz 30 deg2 30 uJy 20” 2013

Apertif 1.4 GHz 8 deg2 50 uJy 15” 2013

MeerKAT

1.4 GHz 1.5 deg2 35 uJy 15” 2013

EVLA 1.4 GHz 0.25 deg2 7 uJy 1.3-45” 2010

EVLA 327 MHz 5 deg2 2 mJy 5-18” 2011

LOFAR 110-240 MHz 50 deg2 1 mJy 5” 2011

EVLA 74 MHz 100 deg2 50 mJy 25-80” 2011

MWA 80-300 MHz 1000 deg2 8 mJy 300” 2011+

LOFAR 15-80 MHz 500 deg2 8 mJy 120” 2011

14

(Only Apertif, EVLA, LOFAR has demonstrated noise perfprmance)

Radio facilities for GW-EM Counterpart Searches: ASCAP

• Australian-lead effort• 36 12-m antennas• Operates at 1.4 GHz• Focal-plane array technology

to give 30 deg2 FoV• 1-hrs, rms~30 uJy (claimed)• 75% of the time given to Key

Science Projects (25% open)– Continuum sky survey 40X

deeper than NVSS– Slow and fast transient

searches

• 2013 delivery (optimistic)

15

Radio facilities for GW-EM Counterpart Searches: Apertif

• Dutch effort• Upgrade of WSRT using

FPAs• 14 25-m antennas• Demonstrated peformance• Operates at 1.4 GHz• 8 deg2 FoV• 1-hrs, rms~50 uJy • 75% of the time will be

given to Key Science Projects (25% open)– Proposals in April 2011

• 2013 operation

16

Radio facilities for GW-EM Counterpart Searches: MeerKAT

• South African-lead effort• 80 12-m antennas• Operates 0.9-1.75 GHz.

Expansion plans 8-14.5 GHz and 0.58-2.56 GHz

• Focal-plane array technology to give 1.5 deg2 FoV

• 1-hrs, rms~35 uJy (claimed)• 75% of the time given to Key

Science Projects (25% open)– Continuum sky survey – Slow and fast transient

searches

• 2013 delivery of 1.4 GHz

17

Radio facilities for GW-EM Counterpart Searches: EVLA

• The 500-lb gorilla of radio astronomy

• 27 25-m antennas• Upgrade project almost

finished. Will deliver order of magnitude increase in continuum sensitivity

• 1-50 GHz + 74 and 327 MHz

• 1-hrs, rms~7 uJy at 1.4 GHz• Responds to external

triggers• Sub-arrays can be used to

image a large error box

18

Radio facilities for GW-EM Counterpart Searches: EVLA

• The 500-lb gorilla of radio astronomy

• 27 25-m antennas• Upgrade project almost

finished. Will deliver order of magnitude increase in continuum sensitivity

• 1-50 GHz + 74 and 327 MHz• 1-hrs, rms~7 uJy at 1.4 GHz• Responds to external

triggers• Sub-arrays can be used to

image a large (irregular) error box

19

Radio facilities for GW-EM Counterpart Searches: LOFAR

• Dutch-lead European project• 36 Dutch stations, 8 Euro

stations• 15-80 MHz & 110-240 MHz• Key Science Projects

– Continuum sky survey– Slow and fast transient

searches

• Real-time pipeline + alert system and external triggers all planned

• RSM will monitor 25% of sky• Million source survey in 2011

20

Radio sky monitor (RSM)

How might we best detect radio signals?Three strategies in order of chance of success

– Afterglow search at late times for off-axis emission• 0.1 to 1 mJy• Timescales of a month• EVLA, ASKAP, MerrKAT, Apertif

– Afterglow search for on-axis event • Bright but rare (i.e. beamed) 1-10 mJy• Timescales of days• EVLA, ASKAP, MerrKAT, Apertif

– Search for prompt signal• 1 mJy to 1 MJy (i.e. highly uncertain)• Low frequency arrays. LOFAR, MWA, electronically steered

in response to GW trigger• Signal will be dispersively delayed

How might we best detect prompt signal?• Prompt signal will suffer

dispersive delay and scattering

• Sources of dispersive delay

– Our Galaxy, IGM, host galaxy and circumburst medium

• Expect DM=1000 pc cm-3, or delays of 13 min at 75 MHz

• Dispersive delay scales as ν-2

• Scattering effects (due to turbulence) are more difficult of estimate.

– 0.1 to 4 s at 75 MHz

– Scattering scales as ν-4.4

DM (pc cm-3)

Lorimer and

Kram

er (2

005)

€

τDM ∝ δν

ν 3

How might we best detect prompt signal?• Prompt signal will suffer

dispersive delay and scattering

• Sources of dispersive delay

– Our Galaxy, IGM, host galaxy and circumburst medium

• Expect DM=1000 pc cm-3, or delays of 13 min at 75 MHz

• Dispersive delay scales as ν-2

• Scattering effects (due to turbulence) are more difficult of estimate.

– 0.1 to 4 s at 75 MHz

– Scattering scales as ν-4.4

DM (pc cm-3)

What follow-up would we want to do?• Panchromatic modeling to

derive real estimates of energy and circumburst density.

• Direct VLBI imaging of the relativistic shock from the afterglow and any radio-emitting detritus from the merger

• Sub-milliarcsecond resolution

• An simple VLBA imaging project. Easier than GRB 030329 (z=0.17)

• Rule of thumb: If LIGO can detect a merger, the VLBA can image it.

GRB 030329z=0.17 (800 Mpc)Pihlstrom et al. (2007)

1 mas at 100 Mpc is 0.5 pc

What follow-up would we want to do?• Panchromatic modeling to

derive real estimates of energy and circumburst density.

• Direct VLBI imaging of the relativistic shock from the afterglow and any radio-emitting detritus from the merger

• Sub-milliarcsecond resolution

• An simple VLBA imaging project. Easier than GRB 030329 (z=0.17)

• Rule of thumb: If LIGO can detect a merger, the VLBA can image it.

GRB 030329z=0.17Pihlstrom et al. (2007)

1 mas at 100 Mpc is 0.5 pc Bietenholz et al. (2003)

SNe 1993J at d=4 Mpc

What follow-up would we want to do?• Panchromatic modeling to

derive real estimates of energy and circumburst density.

• Direct VLBI imaging of the relativistic shock from the afterglow and any radio-emitting detritus from the merger

• Sub-milliarcsecond resolution

• An simple VLBA imaging project. Easier than GRB 030329 (z=0.17)

• Rule of thumb: If LIGO can detect a merger, the VLBA can image it.

GRB 030329z=0.17Pihlstrom et al. (2007)

1 mas at 100 Mpc is 0.5 pc Bietenholz et al. (2003)

What can we be doing today to help field?• Continue to study GW populations

– AM CVn stars– Core collapse (relativistic) SNe– Short-hard bursts

• Characterize the quiescent and transient radio sky to flux densities of 10 uJy

• Develop robust systems to respond to external triggers– Capability to carry out real-time response of radio

telescopes to transients is rare

– Nasu radio transients are an interesting test case. Bright, short lived with poor localization.

Conclusions

• Radio counterpart searches are a powerful tool– Predict a bright signal 1-10 mJy

– Independent of beaming

– Short latency is not needed. (Mañana!)

– False positives are relatively unimportant

• A “bonanza” of new radio facilities is coming on line at just the right times for the next generation GW detectors

28

The future looks brightCome and join the GW-EM adventure

29