Where are the double-degenerate progenitors of type Ia supernovae?

Double white dwarf binaries with merger timescales smaller than the Hubble time and with a total mass near the Chandrasekhar limit (i.e. classical Chandrasekhar population) or with high-mass primaries (i.e. sub-Chandrasekhar population) are potential supernova type Ia (SNIa) progenitors. However, we have not yet unambiguously confirmed the existence of these objects observationally, a fact that has been often used to criticise the relevance of double white dwarfs for producing SNIa. We analyse whether this lack of detections is due to observational effects. To that end we simulate the double white dwarf binary population in the Galaxy and obtain synthetic spectra for the SNIa progenitors. We demonstrate that their identification, based on the detection of Halpha double-lined profiles arising from the two white dwarfs in the synthetic spectra, is extremely challenging due to their intrinsic faintness. This translates into an observational probability of finding double white dwarf SNIa progenitors in the Galaxy of (2.1+-1.0)x10^{-5} and (0.8+-0.4)x10^{-5} for the classical Chandrasekhar and the sub-Chandrasekhar progenitor populations, respectively. Eclipsing double white dwarf SNIa progenitors are found to suffer from the same observational effect. The next generation of large-aperture telescopes are expected to help in increasing the probability for detection by ~1 order of magnitude. However, it is only with forthcoming observations such as those provided by LISA that we expect to unambiguously confirm or disprove the existence of double white dwarf SNIa progenitors and to test their importance for producing SNIa.
Comment: Re-submitted to MNRAS after applying all referee's suggestions