Characterizing the Galactic White Dwarf Binary Population with Sparsely Sampled Radial Velocity Data

Abstract

We present a method to characterize statistically the parameters of a detached binary sample - binary fraction, separation distribution, and mass ratio distribution - using noisy radial-velocity data with as few as two, randomly spaced, epochs per object. To do this, we analyze the distribution of DRVmax, the maximum radial-velocity difference between any two epochs for the same object. At low values, the core of this distribution is dominated by measurement errors, but for large enough samples there is a high-velocity tail that can effectively constrain the parameters of the binary population. We discuss our approach for the case of a population of detached white-dwarf (WD) binaries with separations that are decaying via gravitational wave emission. We derive analytic expressions for the present-day distribution of separations, integrated over the star-formation history of the Galaxy, for parametrized initial WD separation distributions at the end of the common-envelope phase. We use Monte Carlo techniques to produce grids of simulated DRVmax distributions with specific binary population parameters, and the same sampling cadences and radial velocity errors as the observations, and we compare them to the real DRVmax distribution to constrain the properties of the binary population. We illustrate the sensitivity of the method to both the model and the observational parameters. In the particular case of binary white dwarfs, every model population predicts a merger rate per star which can easily be compared to type-Ia supernova rates. In a companion paper, we apply the method to a sample of about 4000 WDs from the Sloan Digital Sky Survey, and we find a merger rate remarkably similar to the rate of Type-Ia supernovae in Milky-Way-like galaxies.