1. Townsley and L. B., 2004, Ap. J., 600, 390 (Theoretical overview) Townsley and L.B., 2005, Ap. J., 628, 395 (Classical Novae) Scannapieco and L.B., 2005, Ap. J., 629, L85 (Type Ia SN Rates) L.B., Townsley, Deloye & Nelemans 2006, Ap. J., 640, 466 (AM CVn) Shen and L. B. 2007, Ap J, 660, 1444 (Stable H/He Burning) L.B., Shen, Weinberg & Nelemans 2007, Ap J., 662, L95 (Faint.Ia SN) Chris Deloye (UCSB=>Northwestern), Gijs Nelemans (U. Nijmegen), Evan Scannapieco (KITP=>ASU), Ken Shen (UCSB), Dean Townsley (UCSB=>U. Chicago) & Nevin Weinberg (KITP=>UCB) Until the final explosion of an accreting WD as a Type Ia supernovae, the brightest manifestations of mass transfer are the thermonuclear ignitions of accreted Hydrogen and Helium. Observing these flashes provides a census of the numbers of such binaries in distant galaxies, allowing for direct comparisons to the Type Ia rates. Explosions on Accreting White Dwarfs

2. Piro ‘05 Accreting White Dwarfs 0.1-1% of white dwarfs are in binaries where accretion occurs, releasing gravitational energy Whereas the nuclear fusion of H=>He or He=>C releases This contrast is further enhanced when the white dwarf stores fuel for > 1000 years and burns it rapidly, making these binaries detectable in distant galaxies during thermonuclear events. Donor star can be H/He or pure He White Dwarf of Carbon/Oxygen

3. M87 in Virgo Some numbers for starters 20 Classical Novae (Hydrogen fuel) per year, implying a white dwarf/main sequence contact binary birthrate (Townsley & LB 2005) of one every 400 years. One Type Ia Supernovae every 250 years, or one in 500 WDs explode! In 10^11 solar masses of old stars (e.g. Elliptical galaxy), two WDs are made per year. The observed rates for thermonuclear events are: Helium novae (Eddington-limited) every ~250 years, one large He explosion every ~5,000 years, and WD-WD mergers every 200 years Predicted rates are:

4. To fill the tidal radius: Accreting White Dwarfs Giving the relation: The mass transfer rate is set by angular momentum losses, typically from gravitational wave emission.

5. Hydrogen Burning is Usually Unstable Townsley & Bildsten 2005 Supersoft Sources: Burn H Stably (van den Heuvel et al 1992), or weakly unstable Cataclysmic Variables: undergo unstable burning, leading to Classical Novae. Whether the mass stays or goes is uncertain Accumulated mass

6. Nova Rates => Population Density The calculated novae recurrence rate as a function of orbital period allows for a census (Townsley & L.B. 2005) of the CV population from the observed nova rate of 20 per year in a 10^11 L_sun, K galaxy (Williams & Shafter 2004), yielding a CV birthrate Mass specific CV birthrate matches the Ia rate in Elliptical galaxies..first time a relative rate comparison is made However, the WD masses in CVs are too low to ignite the C/O in the core, unlikely that they are Type Ia progenitors

7. AM CVn Binaries: Pure Helium Accretors! Found by Humason and Zwicky (‘47) as faint blue stars, spectra by Greenstein & Matthews (‘57) only showed helium lines. Later work found 17 minutes orbital period The accretor is a C/O or O/Ne WD, where the donor is a degenerate Helium WD. Giving an orbital period-donor mass relation, and donor masses ranging from 0.006-0.12 M_sun, 6 Jupiter mass pure He objects!!! RXJ0806 5.35 min V407 Vul 9.49 ES Cet 10.3 AM Cvn 17.1 HP Lib 18.4 CR Boo 24.5 KL Dra 25.0 V803 Cen 26.9 SDSSJ0926 28.3 CP Eri 28.4 2003aw 33.9 SDSSJ1240 37.4 SDSSJ1411 46.0 GP Com 46.5 SDSSJ1552 56.7 CE 315 65.1

8. Optical Spectra are Disk + WD! Emission lines from the accretion disk N/C>100, and N/O=50 due to CNO burning NO Hydrogen, H/He<10^-5 The underlying, nearly featureless continuum is from the reheated white dwarf (Bildsten et al. ‘06) Marsh, Horne and Rosen 1991 The accretion of the helium star exposes material that has completed hydrogen burning

9. GP COM The fate of less than 1 in 2000 white dwarfs in our galactic disk. But none yet seen in other galaxies. These are the brightest Sources for Space-Based Gravitational Wave Detectors, such as LISA!

10. 10 20 30 40 50 60 70 Orbital Period (Minutes) The mass transfer rate and evolution of the binary is driven by loss of angular momentum via gravitational radiation (Faulkner et al ‘72) at the rate And the orbital period increases as the donor loses mass Bildsten et al. ‘06 Time evolution

11. Eclipses in SDSS J0926+3624 Anderson et al (2005) found this object with 28.3 minute orbit, data below is from Marsh et al 2006 from ULTRACAM on WHT g’=19.3

12. Eclipse of the hot white dwarf, the bright spot, and part of the accretion disk. Inclination is 83.1+_0.1 deg. WD mass =0.84+_0.05, and donor mass = 0.03 Marsh et al 2006

13. Helium Burning 10 20 30 40 50 60 70 Orbital Period (Minutes) GP Com CP Eri ES Cet AM CVn CE 315 HP Lib RXJ0806? CR Boo, KL Dra V407 Vul V803 Cen He burning is thermally stable when donor mass is 0.2-0.27, with orbital periods of 2.5-3.5 minutes (Tutukov & Yungelson ‘96). As the accretion rate drops, the burning becomes unstable, and flashes commence. Thick= Cold Donors Thin= Hotter Donors Unstable Helium Burning of Interest

14. L. B., Shen, Weinberg & Nelemans ‘07 Evolution in time Helium Donor Mass For a degenerate (or semi degenerate) donor (dashed lines) there will be many He novae at early times, followed by one last explosive flash with helium masses of 0.03-0.1 solar masses.. For a Helium burning star donor (dotted line; Savonije et al 86; Ergma & Fedorova ‘90), Helium ignition masses >0.2 naturally occur on 0.6 WDs and were studied as double detonations (Nomoto ‘82, Livne ‘90, Woosley et al ‘86, Woosley & Weaver ‘94). L. B., Shen, Weinberg & Nelemans ‘07 Iben & Tutukov ‘89 He Ignition Mass

15. Early in the evolution, many flashes with masses less than 0.01 The last flash has the largest mass (0.03 in these cases), and occurs at 1e-8--1e-7 Msun/year (depending on the accreting WD mass) Last flash mass is sensitive to the binary evolution and ignition mass under changing M_dot.. the 0.8 case nearly ignited at 0.08.. After last flash, the accreted Helium accumulates, so all AM CVns with P_orb> 10 minutes are building He envelopes (Bildsten et al ‘06) Iben & Tutukov ‘89 He Ignition Mass

16. Path to Helium Shell Detonations The radial expansion of the convective region allows the pressure at the base to drop. For low shell masses, this quenches burning. For a massive shell, however, the heating timescale set by nuclear reactions: will become less than the dynamical time, So that the heat cannot escape during the burn, likely triggering a detonation of the helium shell. This condition sets a minimum shell mass.

17. L. B., Shen, Weinberg & Nelemans ‘07 Binary Evolution Naturally Yields Detonations The intersection of the ignition masses with that of the donor yields the hatched region, most of which lie above the dynamical event line ==> For WD masses >0.9, likely outcome is detonation For lower WD masses, the outcome may be less violent, further work needed Must understand the dynamic outcome and nucleosynthetic yields from these low pressure detonations, a new regime.

18. Example Shock goes down (blue arrow) into the C/O and the He detonation (red arrow) moves outward. The shocked C/O under the layer is not ignited. Underlying WD remains Yields at this point in time (0.24 seconds) are 0.012 M_sun of 56 Ni, 0.0071 of 48 Cr, and 0.0076 of 52 Fe. Nevin Weinberg ‘07

19. Thermonuclear Supernova Lightcurves Type Ia result from burning a solar mass of C/O to ~0.6 solar masses of 56Ni (rest burned to Si, Ca, Fe) and ejected at v=10,000 km/sec. This matter would cool by adiabatic expansion, but instead is heated by the radioactive decay chain 56Ni=>56Co=>56Fe Arnett (1982) (also see Pinto & Eastman 2000) showed that the peak in the lightcurve occurs when the radiation diffusion time through the envelope equals the time since explosion, giving The luminosity at peak is set by the instantaneous radioactive decay heating rate.… => can measure the 56Ni mass from Type Ia SN via the peak luminosity, yielding 0.1-1.0 solar masses.

20. The small helium ignition masses (0.02- 0.1) only detonate helium, which leaves the WD at 10,000 km/sec, leading to rapid rise times. The radioactive decays of the fresh 48Cr (1.3 days), 52Fe (0.5 d) and 56Ni (8.8 days) will provide power on this rapid timescale!! L. B., Shen, Weinberg & Nelemans ‘07.Ia Supernovae* *Thanks to Chris Stubbs for the name

21. .Ia Lightcurves courtesy of Daniel Kasen (UCSC) (M_tot,M_Ni) 56Ni Balls 56Ni/Si Balls Delta M_15(B)>3, and sometimes 4 (typical Ia’s have 2 at most)

22. .Ia Supernovae Rates Roelofs et al ‘07 space density implies an AM CVn birthrate If every AM CVn gives a.Ia, their rate would be 2-7 % of the Type Ia rate in an Elliptical Galaxy..Ia Discoveries would reveal distant AM CVns, and may well have been missed in SN surveys due to rapid decline. Volume rate in the nearby universe says that upcoming optical transient surveys with rapid cadences should find many... Daily survey 1/2 sky (LSST) to V=24 would give ~1000.Ia’s per year.

23. Topical Because of New Surveys! Pan-Starrs1 (2008) Sloan Digital Sky Survey (NOW!) Medium deep survey gets 10 per year at -17, and 1 per year at -15 Current survey will find 7 per year at -17, and 0.5 per year at -15. Total duration = 9 months..

24. LSST (2014) Large Synoptic Survey Telescope (LSST) is a proposed ground-based 8.4-meter, 10 square-degree-field telescope that will provide digital imaging of faint astronomical objects across the entire sky every night. Daily survey 1/2 sky would give up to 1000.Ia’s per year. Cerra Pachon, Chile.

25. Bolometric LCs Contardo et al. ‘00, A&A, 359, 876 Subluminous 1991bg “Super” luminous 1991T

26. Type Ia Supernovae Dependence on Galaxy Type and Cosmic Rates There are observed trends in Ia properties with galaxy type (no evidence yet for metallicity effects): 1.Brightest (e.g. 1991T) events occur preferentially in young stellar environments (hence mostly spiral and irregular galaxies) 2.Sub-luminous (and peculiar, eg. 1991bg) Ia’s dramatically prefer old stellar populations.. (Elliptical and S0 Galaxies) 3.Rates track BOTH the stellar mass and the star formation rate These are likely the result of old and young stellar populations and motivated our (Scannapieco & LB, 2005, ApJ, 629, L85) simple explanation for the observed cosmic Ia rate.

27. Rates by Mass Mannucci et al ‘05 Infrared luminosity used to determine the stellar mass Part of the Ia rate tracks the Star formation and is 1/3 the Core Collapse rate Ia Data can be “fit” with one term that depends on mass (confirmed in clusters: Sharon et al ‘06) and another that is 40% of the core collapse rate Roughly one Ia every 400 years for 1 solar mass per year of star formation. SN Rate Dependence on Galaxy Type

28. Assume that the Ia rate tracks the stellar mass and star formation rate as measured by Mannucci et al., then measure the constants from local galaxies to get (Scannapieco & L.B. ‘05) This example shows the outcome when the SFR drops exponentially on a 2 Gyr timescale 80% of the Ia SN over a galaxy’s lifetime come from the prompt contribution (see Oemler & Tinsley 1979). Fe production from Ia’s is 3X that from CC

29. Star Formation Rate

30. Scannapieco & Bildsten 2005 Iron Abundance in Galaxy Clusters

31. Canada-France-Hawaii Telescope SuperNova Legacy Survey (SNLS) Sullivan et al 2006 125 Ia Sne, 0.2<z<0.75 Galaxies identified from the CFHT survey. All Ia’s are spectroscopically confirmed For the clear counterparts (some are ambiguous), the galaxies were classified via colors as vigorous star formers, star- forming, and passive. When SNLS is done, this list should be ~500

32. Scalings with Star Formation Rate Confirmation of the mass specific rate of Mannucci et al for passive galaxies, and confirmation of the Ia rate dependence on star formation rate. Sullivan et al ‘06 1 every 300 years in 10^11

33. CFHT Supernovae Legacy Survey (SNLS) Sullivan et al 2006 Star Forming Galaxies Red=Passive Star Forming-Passive The number of faint (small stretch) Ia’s in spirals is consistent with the old stellar population in the spiral galaxy. The two populations are distinct, but overlapping in their 56Ni production levels. Fainter

34. Sullivan et al ‘06 Their normalization with the SFR is 3 times smaller than Scannapieco and Bildsten ‘05, giving a different evolution with z. Should be resolved, but will take time..... Ia Rate vs. Redshift Blue arrow shows the expected local Ia rate just from the local K- light density

35. Neill et al. 2007 astroph-0701161 CFHT Legacy Survey: Just Fit the Data!

36. Conclusions Classical Novae ignition mass calculations allow us to say that ejected masses are similar to ignition masses.Ia SN expected from AM CVn binaries (which came from double WDs) and should be found in abundance in upcoming surveys. The observed Ia rate depends on both the mass of a galaxy and the star formation rate, allowing for 1. Improved understanding of the Fe abundance in galaxy clusters, 2. Ia rate as a function of redshift, 3. Clear evidence for multiple progenitors Clearly more to learn about thermonuclear events on accreting white dwarfs, especially in the upcoming age of all-sky surveys

37. Las Cumbres Observatory Global Telescope Network A private scientific observatory based in Santa Barbara that owns and operates 2 m Faulkes telescopes in Maui and Australia, and which will: Construct and operate ~6 sites at multiple longitudes with 3x1 meter telescopes for continuous coverage of transients, transits, etc.. Each telescope will have multiple instruments, and be refreshed Construct and operate many (>30) 0.4 meter telescopes around the globe primarily for education and outreach. Remote operations, data management, and instrument construction based in Santa Barbara. Strong connections to UC-Santa Barbara. Tim Brown is LCOGT Scientific Director and Adjunct Professor of Physics at UCSB. The whole facility (both the rings and the smaller telescopes) will be made available to the international scientific community. Fully funded for 25 years of operation. Go to www.lcogt.net