D R . A N N A WAT T S , U N I V E R S I T Y O F A M S T E R D A M

N E U T R O N S TA R A S T E R O S E I S M O L O G Y W I T H M A G N E TA R B U R S T S

D A N I E L A H U P P E N K O T H E N L U C Y H E I L , A L E X A N D E R VA N D E R H O R S T, C H R I S E L E N B A A S , T H I J S VA N P U T T E N , C A R O L I N E D ’ A N G E L O , P H I L U T T L E Y, M I C H I E L VA N D E R K L I S ( U VA ) !C H R Y S S A K O U V E L I O T O U A N D T H E F E R M I G B M M A G N E TA R T E A M

G I A N T F L A R E A S T E R O S E I S M O L O G Y

Giant flare QPOs: Israel et al. 2005, Strohmayer & Watts 2005,6, Watts & Strohmayer 2006

G I A N T F L A R E A S T E R O S E I S M O L O G Y

Giant flare QPOs: Israel et al. 2005, Strohmayer & Watts 2005,6, Watts & Strohmayer 2006

18-1800 Hz Quasi-Periodic Oscillations (QPOs)

N E U T R O N S TA R C O M P O N E N T S

Watts 2013

S E I S M I C V I B R AT I O N M O D E L S

• Coupled magneto-torsional oscillations of crust/core.

• Alfven modes are continua: frequency drifts intrinsic.

• Frequencies depend on mass, radius, superfluidity, crust composition, magnetic field strength and geometry.

• Decay times important.

See for example: Gabler et al. 2013, 14;

Passamonti & Lander 2013, Huppenkothen, Watts & Levin 2014

S M A L L B U R S T A S T E R O S E I S M O L O G Y

Fermi GBM bursts from SGR 0501+4516 (Huppenkothen et al. 2013)

S M A L L B U R S T A S T E R O S E I S M O L O G Y

• QPO searches are complicated by the ‘burst envelope’.

• Traditional method (Monte Carlo simulations of lightcurves) fails in absence of simple functional model for emission.

Huppenkothen et al. 2013

S M A L L B U R S T A S T E R O S E I S M O L O G Y

• We use a Bayesian procedure for modelling the periodogram, assuming a red noise process.

• Method is highly conservative.

S M A L L B U R S T A S T E R O S E I S M O L O G Y

• We use a Bayesian procedure for modelling the periodogram, assuming a red noise process.

• Method is highly conservative.

Unofficially known as the ‘Palin Method’

T H E S G R J 1 5 5 0 - 5 4 1 8 B U R S T S T O R M

Fermi GBM burst storm, January 2009 (Kaneko et al. 2010, van der Horst et al. 2012 )

T H E S G R J 1 5 5 0 - 5 4 1 8 B U R S T S T O R M

Huppenkothen et al. 2014a, b

• Stacking consecutive bursts in most active periods: Significant QPOs (final p < 0.001) in frequency range (93-127 Hz) found in giant flares.

• Similar signals in RXTE data of SGR 1806-20/SGR 1900+14 bursts.

• Global modes excited by burst storms? Excitation threshold?

T H E S G R J 1 5 5 0 - 5 4 1 8 B U R S T S T O R M

Huppenkothen et al. 2014a

• Single burst: Candidate (final p < 0.025) QPO at 260 Hz, different frequency range and broader than giant flare QPOs.

• Core damping? In line with theoretical predictions.

• Signature of trigger mechanism, similar to variability claimed in giant flare peaks?

L I G H T C U R V E M O D E L S

• Is there a simple empirical function that can generate magnetar burst light curves?

• Bayesian hierarchical models: can bursts can be fit as ‘spike cascades’? Yes.

• Testing predictions of simple cascade models for trigger mechanism, including SOC models.

Decompose light curves into superpositions of simple shapes (Huppenkothen et al. in prep).

L I G H T C U R V E M O D E L S

• Is there a simple empirical function that can generate magnetar burst light curves?

• Bayesian hierarchical models: can bursts can be fit as ‘spike cascades’? Yes.

• Testing predictions of simple cascade models for trigger mechanism, including SOC models.

Decompose light curves into superpositions of simple shapes (Huppenkothen et al. in prep).

S U M M A R Y

• Fermi GBM has given us the first robust detections of seismic vibrations in small magnetar bursts

• QPO pipeline ready for future burst storms and the next giant flare

• Can burst storms excite global vibrations?

• Probing time evolution and variability around the trigger point.