One million years until graceland: finding galactic Ia progenitors

AbstractDouble click to add an Abstract

Motivation

The in-spiral and coalescence of double white dwarf (WD) binaries due to gravitational wave emission has several possible outcomes, all of them of the highest astronomical interest. When the combined mass exceeds the Chandrasekhar limit ($$M_{\rm Ch}$$), the merger could lead to a Type Ia Supernova (SN) (Iben et al., 1984; Webbink, 1984), although accretion induced collapse to a neutron star is also a possibility (Saio et al., 1998; Schwab et al., 2016). Recent work shows that the delay time distribution of SN Ia behaves like 1/t for t $$\gtrsim$$ 1 Gyr, exactly as predicted by a population of binaries that merge through gravitational wave emission (Maoz et al., 2014). Sub-Chandrasekhar WD mergers have been linked to a new class of fast, faint transients with low ejected mass and unusual nucleosynthetic signatures (Bildsten 2007, Shen 2010), like SN2002bj (Poznanski 2009) and SN2005E (Perets 2010) or even possibly normal SN Ia (Kerkwijk 2010, Sim 2010). Although several binary WD systems have been discovered in the last years (e.g., Gianninas et al., 2015), the vast majority of the known pre-mergers have low mass He WD primaries, and combined masses well below $$M_{\rm Ch}$$. Additionally, current empirical constraints on the merger rate are highly uncertain due to the very small number of known short period systems (Brown 2016).

Basic Idea

Compact binary WDs have large orbital velocities, with $$v_{\rm orb}\simeq 500\,{\rm km\,s}^{-1}(P_{\rm orb}/15\,{\rm min})^{-1/3}$$ for an equal-mass $$0.6\,M_{\odot}$$ WD binary. Such binaries should exhibit large radial velocities (RVs) that can be measured even in low resolution wide-field spectroscopic surveys. WD binaries with orbital periods of minutes to hours can thus be found with follow-up observations of targets with the largest RV variations.

The odds of detecting WD binaries at short orbital periods are good. If all type Ias are produced by C/O WD mergers, we expect a WD binary to merge within the Milky Way every $$\sim 10^{2}\,{\rm yr}$$. The orbital decay time is $$200\,{\rm Myr}$$ at an orbital period of 2 hours, therefore we estimate $$\sim 2$$ million WDs with orbital periods under two hours within the galaxy. WDs at short orbital periods (under 15 minutes) may be tidally heated to surface temperatures $$T{\rm eff}\gtrsim 20,000$$K (Fuller et al., 2013), brightening them significantly and increasing the odds of detecting these systems in magnitude-limited surveys.

The SWARMS survey (Badenes et al., 2009; Mullally et al., 2009) was started in 2009 to identify pre-merging binary WDs in the large ($$\sim$$15,000 objects) database of WD spectra accumulated by the Sloan Digital Sky Survey. SWARMS uses the $$\sim$$15 minute sub-exposures to look for RV shifts (Figure 1). The typical time lag between sub-exposures ($$\sim$$ half an hour to a few hours) is comparable to the orbital periods of pre-merging WDs, and the expected RV shifts ($$\gtrsim$$ 100 km s$${}^{−1}$$) can be easily detected at the spectral resolution of SDSS, putting SWARMS in a good position to identify most of the pre-merging systems in this data base (see Badenes et al. 2009 for details). Unfortunately, followup of the SWARMS