The binary fraction of stars in dwarf galaxies: the case of Leo II

We combine precision radial velocity data from four different published works of the stars in the Leo II dwarf spheroidal galaxy. This yields a dataset that spans 19 years, has 14 different epochs of observation, and contains 372 unique red giant branch stars, 196 of which have repeat observations. Using this multi-epoch dataset, we constrain the binary fraction for Leo II. We generate a suite of Monte Carlo simulations that test different binary fractions using Bayesian analysis and determine that the binary fraction for Leo II ranges from $0.30^{+0.09}_{-0.10}$ to $0.34^{+0.11}_{-0.11}$, depending on the distributions of binary orbital parameters assumed. This value is smaller than what has been found for the solar neighborhood (~0.4-0.6) but falls within the wide range of values that have been inferred for other dwarf spheroidals (0.14-0.69). The distribution of orbital periods has the greatest impact on the binary fraction results. If the fraction we find in Leo II is present in low-mass ultra-faints, it can artificially inflate the velocity dispersion of those systems and cause them to appear more dark matter rich than in actuality. For a galaxy with an intrinsic dispersion of 1 km/s and an observational sample of 100 stars, the dispersion can be increased by a factor of 1.5-2 for Leo II-like binary fractions or by a factor of 3 for binary fractions on the higher end of what has been seen in other dwarf spheroidals.