Pulsating WD Paula Szkody, Anjum Mukadam, Boris Gaensicke, Arne Henden + AAVSO, Atsuko Nitta, Ed Sion, Dean Townsley Wild Stars March 16, 2009 Tantalizing Hints from CVs with Pulsating WDs

Why study WD pulsators ? Pulsations probe inner stellar structure uniform density; pulsations penetrate the inner 99% of white dwarf Asteroseismology allows determination of:  Mass (total and H layer)  Rotation rate  Core Composition  Cooling rate

Why study accreting WD pulsators ? Find effects of accretion on pulsations  Heating due to accretion  Increase in angular momentum  Changing surface composition  Changing stellar mass

White dwarfs show non-radial g-modes on account of their high gravity Periods of 100s to 1000s These modes are characterized by quantum numbers (k,l,m) similar to atomic orbitals k is #nodes between surface and center l is # nodes on surface (borders between hot and cool zones) m is # on surface that pass through pole of rotation axis

Modes of Pulsation l = 1 m = 1 m = 0 l = 2 m = 1 m = 0 Stellar Surface see

Ideal short term seismology ● Find a star with many (~10) pulsation modes ● Identify the quantum numbers for each mode ● Use WD pulsation models to get a unique fit to the observed periods Model fit Temperature, stellar mass H/He layer masses Model luminosity Distance to star Multiplet structure Magnetic field or rotation Model core composition 12 C, 16 O nuclear reaction rate in RG stage

3 major questions: 1) where is the instability strip for accreting pulsators? 2) what are the affects of accretion and outbursts? 3) is there any hope of pulling out mass, spin, etc?

Instability Strips for Non-Interacting White Dwarf Pulsators ZZ Ceti K 25000K Our question: where is the instabilty strip for accreting pulsators?

Step 1: Find some Accreting Pulsators Cataclysmic Variables at quiescence with very short ~80-90 min orbital periods so accretion rate at a minimum M  /yr 90% of the optical light comes from the white dwarf So it’s possible to detect white dwarf pulsations! 13 known accreting pulsators to date Need:

It’s a short history ( ): GW Lib ,370,230Warner & van Zyl 1998 SDSSJ ,347I. Woudt & Warner 2004 SDSSJ ,475,330II. Warner & Woudt 2004 SDSSJ ,335,260II. Warner & Woudt 2004 V455 And (HS) Araujo-Betancor et al PQ And ,1283 Vanlandingham et al,Patterson et al 2005 REJ ,1236,668Patterson et al 2005 SDSSJ IV. Nilsson et al 2006 SDSSJ IV. Gansicke et al 2006 SDSSJ ,647V. Mukadam et al 2007 SDSSJ IV. Mukadam et al 2007 SDSSJ ,660,500IV. Patterson et al 2008 SDSS V. Pavlenko 2008 P orb (min) V (mag) P pulse (sec) Ref

signature of pulsating, accreting WDs

Step 2: Get UV data Ly  region allows best T determination Pulse amplitudes 6-17 x larger in UV

UV time-series spectroscopy Space Telescope Imaging Spectrograph (STIS) A with 1A resolution, time-tag HST Advanced Camera for Surveys (ACS) Solar Blind Channel (SBC) Å 6-40A resolution, 1 min integrations Co-added Spectrum & Light Curve HST Programs 2002/2008 UV data on 9 so far until SM4!

Ground data are necessary: We need to know the optical state (to make sure object is at quiescence - HST insists) Simultaneous UV & opt pulsation amplitudes are needed to determine the mode

HST Obs AAVSO + help  HST picks week months in advance  AAVSO issues alert/web page ~ 2 wks-1 month before HST  AAVSO monitors  HST gives date/time 1 week prior to obs  ground obs 24 hr prior t akes halt off  HST observes with ground support GW Lib Jan 17, 2002 STIS V455 And Oct 24, 2002 STIS snap SDSS0131 Jun 18, 2005 SBC SDSS2205 May 23, 2005 SBC SDSS1610 June 30, 2005 SBC PQ And Sep SBC SDSS0745 Nov 1, 2007 SBC SDSS0919 Nov 14, 2007 SBC SDSS1339 Jan 25, 2008 SBC Object Date Instrument

fits with 2 component WD: 63% at 13,300K, 37% at 17,100K STIS on GW Lib

16,500 WD model Accretion disk SBC Fit to HST spectrum of SDSSJ

10,000K WD 12,000K WD 15,000K WD HST data SDSS ,500K WD SDSS data UV + opt with WD models

V455 And = HS ,000K 20,000K 12,000K 11,000K STIS + Calar Alto data on

HST data on PQ And SBC SBC+opt+WD models ,000K

Temperatures from HST UV spectroscopy Spectral Fits by Boris Gansicke Uncertainties ~ 1000K GW LibJan17, 2002STIS15,400 V455 AndOct 24, 2002STIS (snap)10,500 SDSS May 23,2005SBC15,000 SDSS Jun 18, 2005SBC14,500 SDSS Jun 30, 2005SBC14,500 PQ AndSep 13, 2007SBC12,000 SDSS Nov 1, 2007SBC16,500 SDSS Nov 14, 2007SBC13,500 SDSS Jan 25, 2008SBC12,500 Object Date Instrument Temp (K) SDSS Littlefair et al 2007 eclipse phot 11,000

Latest ZZ Cet instability strip (Gianninas, Bergeron, Fontaine, 2009) High mass DA WDs can get up to ~14000K log g T eff (K) Is the T difference a Mass difference?

(Arras, Townsley & Bildsten 2006, ApJ, 643, L119) Helium Abundance < 0.38 Instability Strip like ZZ Ceti Helium abundance > 0.38 Additional instability 15000K Helium Abundance > 0.48 Strips merge into a wide strip K Is the difference based on mass and evolution? Is it a Composition Difference?

Past HST observations show CVs with WDs T’s that should place them in the instability strip but they don’t show pulsations Of 7 WZ Sge stars with WD = K, no obvious pulsations in EG Cnc, LL And, AL Com, HV Vir, WX Cet, VY Aqr, EF Peg Is this lack of data or lack of pulsations? Another part of the enigma:

STIS data on EG Cnc T=12,300K WD Szkody, Gansicke, Sion & Howell, 2002, ApJ, 574, 950

Optical data on SDSS telescope and cadence matter 1.2m, 70s int 3.6m, 20s int

Further Clues come from the pulsation analysis... The first 4 observations were classic cases: UV periods same as optical UV/opt amplitude ~ 6-17 as predicted by theory

GW Lib HST-STIS Szkody, Gansicke, Howell, Sion 2002, ApJ Orb 1 Orb 2 Orb 4 Orb 3

Szkody et al. (2007, ApJ, 658,1188) P=214s P=608s P=576s in 2005

DA model for log g=8 by Atsuko Nitta with (Szkody et al. ApJ 658, 2007) T eff = 14,500 T eff = 12,500 UV+optical data allow fit to mode SDSS s pulsation T from spectrum = 14,500K

Then it got complicated: PQ And showed optical but no UV pulse SDSSJ0745 and others showed no pulse in opt or UV!

1285s 2337s UV/Opt amplitude < 1! l=3?

GW Lib650,370, SDSS ,213, SDSS ,221 6 SDSS ,475, PQ And c1263none<1 SDSS none SDSS none SDSS1339 c642230,210? V455 And c419,336,310No obs SDSS none REJ ,1236, 654,582 none<2 Object Opt Periods (s) UV Periods (s) UV/opt amp Summary of Pulsations

Optical Data prior to HST on Long Period Pulsator SDSS Light curvesFourier transforms

But in 2007: Orb P

SDSS Oct 31 Oct 1 Nov 2 Nov 6 peaks/ 7200 sec = pulsation1.5 peaks/ 7200s = orbital

This accreting pulsator stopped pulsating between ! No respectable ZZ Ceti stops pulsating! [Anjum] The Catalina Sky Survey detected an outburst of SDSS in Oct did this matter? BUT

Unknown SOB occurred or Accreting pulsators have changing temperature, mass, angular momentum & surface composition ??? We need long-term observations for each pulsator! Possible Reasons for non- detection of pulsations:

GW Lib P = 650, 376, 236s Long monitoring shows that amplitudes can change (Van Zyl et al 2004, MNRAS)

Patrick Woudt SAAO long term monitoring of pulsator SDSS0131 From Szkody et al. ApJ 658, 1188, 2007 HST data June 2005 showed P=214s 214s 600s 335s

3 major questions: 1) where is the instability strip for accreting pulsators? 2) what are the affects of accretion and outbursts? 3) is there any hope of pulling out mass, spin, etc?

Known Outbursts of Accreting Pulsators: PQ And (1938, 1967, 1988) GW Lib (1983, 2007) V455 And (2007) REJ (1994) SDSS (2006) SDSS (2006)

Pulsation Period: Means of measuring T eff ? Mukadam et al. 2005

AAVSO data plotted by Matt Templeton Apr12 07 more GALEX May, June 2008 May,June 07 GALEX WIYN,APO May29,June 1 (387+/-5s? 19 mma) ULTRACAM (Copperwheat et al. 2009) GW Lib outburst

13 months past outburst

A new periodicity is apparent 1 yr after outburst but not seen in UV ??? Before outburst: P=650, 370, 230s After outburst: P=1145+/-3 s Mar-June 2008 Copperwheat et al MNRAS: Mar-Apr 1154+/-10s June s 1145s

CTIO 4m spectrum: Tom Harrison Apr 2008 Boris fit: T wd = 23,000K, log g=8.46 Quiescent WD T=15,000K V~16.3

V V455 And AAVSO data plotted by Matt Templeton Sept 07 V455 And

Pre SOB Post SOB spinpulse? 1 yr past outburst:

2 systems like this now known Phot P ~ 43 min Spect P = 85 min Spectra look like pulsators Pre-outburst photometry

Pavlenko Dec 2008 astro-ph Feb 06 Oct 06 Feb 07 Jan 08May 08

Oct 06 Feb 07 Pavelenko astro- ph 2008 Jan 08 Pulse P = 756s Did this system move INTO the instability strip after outburst? No pulsations in pre-outburst data

Past HST data Model 10,500K WD Maybe we can solve this in the next HST cycle with T, mass measurements 12500K COS time for GW Lib (high T) and V455 And (low T)

3 major questions: 1) where is the instability strip for accreting pulsators? 2) what are the affects of accretion and outbursts? 3) is there any hope of pulling out mass, spin, etc?

Coordinated Campaign on SDSS May 2007

Nice power spectrum but no additional pulsation periods!

If triplet is due to WD rotation : P=4.8days! van Zyl et al (2004) found similar result for GW Lib: splitting implied rotation P =7.3d! other causes of splitting - magnetic field?

3 major questions: 1) where is the instability strip for accreting pulsators? 11, ,000K 2) what are the affects of accretion and outbursts? pulsations disappear, mode splitting 3) is there any hope of pulling out mass, spin, etc? difficult - need many more modes and data

What remains to be done to make progress: Observationally : find the rest of the pulsators (SDSSVI,VII, lit) fill in the temperatures of the instability strip monitor GW Lib and V455 And as WDs return to T quies have more campaigns to look for multiplet structure Theoretically: find a way to make pulsations disappear explain fine structure in pulse include WD spin, composition in models