Sarah Burke Spolaor

Jet Propulsion Laboratory, California Institute of Technology

Gravitational Wave Detection with Pulsar Timing Arrays:Status and Prospects

© 2013 California Institute of Technology, Government Sponsorship Acknowledged

Millisecond pulsars

Spinning up to ~700 times per

second

“Timing Residuals”Model pulsar Observe Correct?

Mod

el

– A

ctu

al

Arr

ival

Ph

ase

(m

s)

Time (relative MJD)

Figure of Merit: RMS scatter of

residuals.

BEST: <50nsWORST: few ms

Fit for known effects

Example pulsar modelP

SR

J0437-4

715

Also referenced:

• JPL Planetary ephemeris

• TAI international atomic time standard

Pulsar

Earth

Jenet et al. (2004)

Pulsar Timing Array

• Monopolar signature?• Atomic time standards (Hobbs et

al. 2012)• Telescope issues

• Dipolar signature?• Planetary ephemeris errors

(Champion et al. 2010)

• Quadrupolar signature?• Gravitational waves

GW SPECTRAL BAND:Observing cadence Experiment length

~2 weeks to >10 yearsnHz – mHz

GW Spectrum

Adapted from Yardley et al. (2009)

log

[d

imen

sion

less

GW

str

ain

]

Stochastic SMBH Binary Background

UNTIL RECENTLY:“Working on our sensitivity”

CURRENTLY & UPCOMING: Meaningful upper limits

+Detection

GW BackgroundN

orm

ali

zed

Dis

trib

uti

on

Strain Amplitude at f = (1 year)-1

ALL MODELS

Fiducial models

Low-mass BCG

High-mass BCG

Adapted from Sesana et al (2013)

GW Background

Van Haasteren et al. 2013

Shannon et al. (accepted to

Science)

Norm

ali

zed

Dis

trib

uti

on

Strain Amplitude at f = (1 year)-1

ALL MODELS

Fiducial models

Low-mass BCG

High-mass BCG

Rules out standard Millennium Simulation

binary presecription to 50% confidence

Sensitivity scaling law

S/N

Number of pulsars

Average residual RMS Number of observations

Length of experiment

Scaling law from Siemens et al. (2013)

b = 13/3 for SMBH binary background

Recent Sensitivity Improvements:

Gaussian & Non-stationary Noise

Recent Sensitivity Improvements: Detection

Algorithms• Coherently seek correlations using all pulsars

• More sensitive statistical analysis

• Resolved sources:

• Corbin+Cornish10; Finn+Lommen+10; Lee+11; Ellis+12; Boyle+Pen12; Mingarelli+12; Ellis13 …

• Sky localization (~2000 deg2; Ellis 2013)

• Parameter estimation (M, e, D, P …)

• Measuring Spin-orbit Precession

Recent Sensitivity Improvements: Detection

Algorithms

Incoherent spectral analysis

(Yardley+09)

Bayesian inference(Ellis et al. in prep)

Thanks to J. Ellis for figure

Yardley et al. (2009) data set: two algorithms

Recent Sensitivity Improvements: Detection

Algorithms• Coherently seek correlations using all pulsars

• More sensitive statistical analysis

• GW Backgrounds:

• van Haasteren+11; Demorest+12; Shannon et al (accepted)

• IPTA data challenge (12 distinct submissions, paper in prep)

Recent Sensitivity Improvements:

International Pulsar Timing Array

• Nanohertz Observatory for Gravitational Waves (NANOGrav; North America)

• European Pulsar Timing Array (Europe)

• Parkes Pulsar Timing Array (Australia)

http://www.ipta4gw.org

Recent Sensitivity Improvements:

International Pulsar Timing Array

• DOUBLE number of pulsars [~40 total]

• LONGER data sets [up to 30 years]

• LOWEST RMS RESIDUALS pulsars [many under 500ns]

• LARGE NUMBER OF DATA POINTS

S/N

Number of pulsars

Average residual RMS Number of observations

Length of experiment

Scaling law from Siemens et al. (2013)

b = 13/3 for SMBH binary background

100 Pulsars10 yr per pulsarCoherentOptimistic timing precision

The Future:Resolved SMBH Binaries

z = 0.001

z = 0.01

z = 0.1

Optimistic Future timing array

with Square Kilometre Array

Burke-Spolaor (2013; CQG Special issue on Pulsar Timing Arrays)

Confusion limit?(Boyle & Pen 2012)

2e9Msun atInternational Timing Array 2014

+ Ellis+12 Bayesian algorithm

Yardley et al. (2010)

The Future: GW Background

Shannon et al. (submitted)

Square Kilometre Array100 pulsars, RMS < 100ns,

for 10 years

Norm

ali

zed

Dis

trib

uti

on

Strain Amplitude at f = (1 year)-1

ALL MODELS

Fiducial models

Low-mass BCG

High-mass BCG

IPTA est.

The Future: GW Background

With three new

pulsar discoveri

es per year

Continuing without improvem

ent

Only NANOGrav considered here (Siemens et al. 2013)

Summary• Galactic-scale gravitational wave observatory

• Supermassive black hole binaries anticipated first detection: Individual/Stochastic Background

• Gravitational waves in ~9 years WITHOUT improvements.• IPTA formation• Enhanced algorithms and more pulsars• Improved instrumentation + understanding of

“detector” (pulsar)

• Timing Array science not covered:• Multi-messenger targets• Strongest observational limits on cosmic string tension• Testing alternate theories of gravity• Detecting trans-Neptunian objects• Spacecraft naviation with timing arrays

Grab-bag:Alternative gravity

theoriesLee+08

Where to look?Burke-Spolaor (2013; CQG Special issue)

References: Comerford+09, Liu+10, Shen+11, Komossa+03, Fabbiano+11, Graham04, Milosavljevic+Phinney05, Sesana+11, Tanaka+12, Eracleous+11, Burke-Spolaor11, Gower82, Volonteri+08, and more

Red: not yet confirmed

Grab-bag: Astrophysics with GW

limits• 3C66B (Sudou+03, Jenet+04)

• 1.06 year orbit (Pgw = ½ year)

• Total mass > 1010 Msun

Simulated 3C66B signal… Actually saw…