Presentation #102.366 in the session Poster Session.
We present a parametric model of companion frequency as a function of mass ratio (from < 0.001 to > 0.1), orbital separation (< 0.3 to > 300 AU), and host star mass (0.2-3.0 Msun). Such a model can help guide predictive theories of star and planet formation, as well as improve predicted yields of very low mass companion surveys, explicitly taking into account contributions from both planet and brown dwarf companion populations. We fitted 36 point estimates of companion frequency using data from a wide range of survey techniques including radial velocity and direct imaging. We assume that: a) the companion mass functions are independent of orbital separation; and b) brown dwarf (BD) companions follow the same orbital distributions as stellar multiples. The gas giant planet orbital distribution can be fitted as a log-normal with a peak between 3-6 AU and recovers a power-law planet mass ratio distribution dNplanet /dq ~ q-1.3 for host stars M through A type. We also recover a relatively flat mass ratio distribution (dNBD /dq ~ q-0.4) for brown dwarf companions;. We find a single normalization constant for the planet populations, and another for the BD population. Yet because the peak in the distribution of binary companions depends on host star mass, the local minimum in the combined companion mass ratio distribution varies with host star mass and orbital separation. We explore the use of this model to interpret results from new direct imaging surveys including the ESO SPHERE SHINE GTO Program (PI: J.L. Beuzit), as well as the ESO BEAST Large Program (PI: M. Janson). The model assesses the probability that an object of a given q and orbital separation was drawn from a “planet-like” or “binary-like” population, and is consistent with all observational trends recorded to date.