Accrete, accrete, accrete Bang! (and repeat): The remarkable recurrent novae

Transcription

1 Accrete, accrete, accrete Bang! (and repeat): The remarkable recurrent novae Ma= Darnley Liverpool John Moores University Co-conspirators: Mike Bode (LJMU), MarGn Henze (CSIC-IEEC), Rebekah Hounsell (Urbana-Champaign), Mariko Kato (Keio), Tim O Brien (Manchester), Valerio Ribeiro (Radboud), Allen ShaRer (SDSU), Mike Shara (AMNH), and Steve Williams (Lancaster)

2 Novae: Vital StaGsGcs Primary: White Dwarf 0.5 ~1.4 Solar Masses (Chandrasekhar mass) CO or ONe(Mg) Plus He and H layers Secondary: `Late Type MS star Typically Solar-ish composigon Orbital period hrs Roche lobe filling Mass AccreGon: Solar material Hydrogen rich AccreGon rate: 10-9 Solar masses / year Luminosity: Quiescence: ~ Solar Luminosity ErupGon: ~10 4 Solar Eddington luminosity Ejected Mass: Solar Masses ComposiGon: depleted H, enhanced He/N Possible Ne enhancement is ONe WD Recurrence Period: 1,000 1,000,000 years 1,000s of erupgons per system Recurrent Novae: >1 observed erupgon Secondary: Evolved Orbital period: >1 day High accregon rate (up to 10-7 M sun / yr High mass WD (>1.2 M sun )

3 Nova mechanism

4 Classical vs Recurrent Novae Classical Novae Only one observed erupgon Recurrence Gmescales few thousand million years (predicted) Secondaries main sequence stars Low mass transfer/accregon rates Range of WD masses Range of evolugon speeds Range of ejecgon velociges Mix of Fe II and He/N spectra 400 GalacGc systems (1000 M31) WD mass decreasing Recurrent Novae >1 observed erupgon Recurrence Gmescales years Secondaries sub-giants (U Sco) or red giants (RS Oph) High mass transfer/accregon rates High WD Masses Rapid erupgon evolugon High ejecgon velociges He/N spectra 10 GalacGc systems (16 M31, 3 LMC) WD mass increasing è SN Ia (?)

5 Classical vs Recurrent Novae Classical Novae Only one observed erupgon Recurrence Gmescales few thousand million years (predicted) The problema6c systems Secondaries main sequence stars Low mass transfer/accregon rates Range of WD masses Recurrent Novae >1 observed erupgon Recurrence Gmescales years Secondaries sub-giants (U Sco) or red giants (RS Oph) GK Per classical nova hosgng a sub-giant secondary T Pyx recurrent nova with a main High sequence mass transfer/accregon secondary, rates low mass WD, slowly evolving erupgons, High WD Fe Masses II spectrum Range of evolugon Plus speeds others! Rapid erupgon evolugon Range of ejecgon velociges High ejecgon velociges Mix of Fe II and He/N spectra He/N spectra 400 GalacGc systems (1000 M31) 10 GalacGc systems (16 M31, 3 LMC) WD mass decreasing WD mass increasing è SN Ia (?)

6 The Liverpool Telescope The World s Largest Fully Robo)c Telescope

7 M31N b Bode & Darnley et al The first extragalacgc nova to be observed in opgcal, UV and X-rays, and spectroscopically confirmed IniGally thought to be recurrent due to posigonal coincidence with a X-ray observagons and opgcal spectra consistent with recurrent nova HST archival data revealed progenitor system red giant secondary First recovery of a nova progenitor beyond the Milky Way

8 M31N b Bode & Darnley et al The first extragalacgc nova to be observed in opgcal, UV and X-rays, and spectroscopically confirmed IniGally thought to be recurrent due to posigonal coincidence with a X-ray observagons and opgcal spectra consistent with recurrent nova HST archival data revealed progenitor system red giant secondary First recovery of a nova progenitor beyond the Milky Way

9 GalacGc Progenitors Darnley The problem with GalacGc nova studies is the distance... the problem with extragalacgc studies is the distance(!) Can we carry out a stagcally significant survey of extragalacgc progenitors AssumpGon: the colour & magnitude of a quiescent nova can be used to determine secondary type Colour-mag posigon is determined by the secondary accregon disk shirs emission blue-wards and luminosity upwards Blue MS-novae (CNe); Green SGnovae; Red RG-novae

10 GalacGc Progenitors Darnley The problem with GalacGc nova studies is the distance... the problem with extragalacgc studies is the distance(!) Can we carry out a stagcally significant survey of extragalacgc progenitors AssumpGon: the colour & magnitude of a quiescent nova can be used to determine secondary type Colour-mag posigon is determined by the secondary accregon disk shirs emission blue-wards and luminosity upwards Blue MS-novae (CNe); Green SGnovae; Red RG-novae

11 ExtragalacGc novae

12 ExtragalacGc Progenitors Darnley+ 2013

13 M31 Nova Survey ShaRer & Darnley et al. (2011), Williams & Darnley et al. (2014, 2016) Extensive follow-up survey of M31 novae: August 2006?? (Sala in prep) Lead faciliges: Liverpool Telescope (phot +spec), HET (spec), HST (archive) Photometric, spectroscopic, astrometric follow-up survey of almost 100 novae (3x the spectroscopically confirmed sample) As with M31N b, uglising the HST archive for quiescent detecgon

14 M31 Quiescent Nova Survey Williams & Darnley et al. (2014)

15 M31 Quiescent Nova Survey Williams & Darnley et al. (2016) Progenitor Survey Highlights 30±12 % of M31 nova erupgons occur in RGnova systems The tradigonal GalacGc proporgon is ~3 % Order of magnitude increase in the RG-nova populagon size same affect to their SN Ia contribugon(?) RG-novae are associated with the disk (young) stellar populagon of M31 Result is consistent with all the RG-novae being disk novae No nova SN Ia contribugon from novae in early-type galaxies(?)

16 RNe / RG-novae è SNe Ia (?) Many models now show RN WD mass is growing with Gme He-flashes on high mass WDs may not cause significant mass loss High mass WD and high accregon rate drive shorter recurrence Gmes Short recurrence Gmes drive WD mass growth rates PredicGons limit recurrence periods to 50 days as approach Chandrasekhar limit Minor quesgon about the WD composigon CO vs. ONe GalacGcally RG-nova: RS Oph 20 yrs SG-nova: U Sco 10 yrs LMC LMCN a 6 years (as of 2016) M31 M31N c 5 years But wouldn't it be nice to find something more extreme?! A system much closer to the 50 day limit? But we need to find the extreme system(s)

17 M31N a 2008 M31 transient discovered 2009 ErupGon announced (PTF; Tang+ 2014) 2010 ErupGon recovered (Henze+ 2015) 2011 Erupted again 2012 Spectroscopically confirmed 2013 iptf discover erupgon 2014 LT detects predicted erupgon (Darnley+ 2015) Transient X-ray detecgons: 1992, 1993, and 2001 At least one opgcal erupgon in 1960s

18 M31N a: The missing 2010 erupgon Henze & Darnley et al

19 M31N a: The recurrence period Henze & Darnley et al The OpGcal ErupGons 26 Dec Dec Nov Oct Oct Nov Oct 2014 Extrapolated from X-ray detecgons 5 Feb 1992 (ROSAT) 11 Jan 1993 (ROSAT) 8 Sep 2001 (XMM)

20 M31N a: Recurrence period opgcal erupgons Henze & Darnley et al t rec = 351 ± 13 days

21 M31N a: The recurrence period Henze & Darnley et al The OpGcal ErupGons 26 Dec Dec Nov Oct Oct Nov Oct 2014 Extrapolated from X-ray detecgons 5 Feb 1992 (ROSAT) 11 Jan 1993 (ROSAT) 8 Sep 2001 (XMM)

22 M31N a: Recurrence period vs. X-ray detecgons Henze & Darnley et al. 2015

23 M31N a: Recurrence period vs. X-ray detecgons Henze & Darnley et al. 2015

24 M31N a: A recurrence period of 6 months? Henze & Darnley et al t rec = 175 ± 11 days

25 M31N a: Where are the missing erupgons? Henze & Darnley et al. 2015

26 M31N a: The recurrence period Darnley+ in prep The OpGcal ErupGons 26 Dec Dec Nov Oct Oct Nov Oct Aug 2015 Extrapolated from X-ray detecgons 5 Feb 1992 (ROSAT) 11 Jan 1993 (ROSAT) 8 Sep 2001 (XMM)

27 M31N a: 2015 erupgon r -band Darnley+ in prep

28 M31N a: erupgon r -band Darnley+ in prep

29 M31N a: Always the same? Darnley+ in prep

30 M31N a: 2015 erupgon opgcal spectra Darnley+ in prep

31 M31N a: 2014 vs 2015 Hα Development Darnley+ in prep

32 M31N a: 2014 vs 2015 Hα Development Darnley+ in prep

33 M31N a: SupersoR X-ray source light curve Darnley+ in prep No detecgon of SSS following 2012 erupgon First X-ray detecgon followed 2013 erupgon Turn-on detected by SwiR 5.9 ± 0.5 days post erupgon BB fit to X-ray spectra gives kt = 120 ± 5 ev (~1.4 million K) Count rate declined at day 16: turn-off 18.4 ± 0.5 days

34 M31N a: Hubble Space Telescope Campaign

35 M31N a: Hubble Space Telescope Campaign Darnley+ in prep Cycle 23 Proposal Successful (yay) Rapid STIS NUV and FUV (low res spectroscopy; Angstroms) DToO STIS FUV-MAMA 4 orbits, 1 visit DToO STIS NUV-MAMA 4 orbits, split over 2 visits 4 WFC3/UVIS visits (F225W, F275W, F336W, F475W, F814W) Visit 1 3 orbits - ~day 14 Visit 2 3 orbits - ~day 21 Visit 3 3 orbits - ~day 28 Visit 4 3 orbits - ~day 35 (~quiescence)

36 M31N a: HST photometry (F336W U-band) Darnley+ in prep

37 M31N a: HST photometry Darnley+ in prep

38 Quiescent photometry Darnley+ 2014

39 M31N a: HST PHAT photometry Darnley+ in prep

40 M31N a: The remnant?!?!(?!) Darnley+ 2015

41 M31N a: The remnant s spectrum! Darnley+ 2015

42 Models of M31N a WD mass 1.35M, dm/dt > M /yr (Kato+ 2014; based on 2013 data) ErupGons consistent with models of systems with >1.377M & 1.358M WDs (Henze+/Kato+ 2015; 2013/14 data) AccreGon rate > 10-7 M / year Mass accumulagon efficiency 0.64 (Kato+ 2015) Timescale to reach Chandrasekhar mass < 0.6 Myr (Tang+ 2014)

43 Conclusions / ImplicaGons / QuesGons A recurrent nova with an unprecedented 1 year (six months?) inter-erupgon Gme M31N a is the prime SN Ia SD-channel pre-explosion progenitor candidate Are such systems rare or common? Secondary type (RG vs. SG) sgll relagvely uncertain accregon mechanism? White dwarf composigon (CO vs. ONe) unknown can we ever know this? 1 sigma confidence periods for 2016 erupgons: 26 th Feb 19 th Apr & 21 st Aug 13 th Oct DetecGon and follow-up campaigns for the 2016 erupgon(s) coming together including LT, LCOGT, XMM (all confirmed), plus HST, Swi., GTC, Gemini, Keck, LBT + more ground-based