Finding benchmark brown dwarfs to probe the IMF as a function of time

Using a simulated disk brown dwarf (BD) population, we find that new large area infrared surveys are expected to identify enough BDs covering wide enough mass--age ranges to potentially measure the mass function down to ~0.03Mo, and the BD formation history out to 10 Gyr, at a level capable of establishing if BD formation follows star formation. We suggest these capabilities are best realised by spectroscopic calibration of BD properties (Teff, g and [M/H]) which, when combined with a measured luminosity and an evolutionary model can give BD mass and age relatively independent of BD atmosphere models. Such calibration requires an empirical understanding of how BD spectra are affected by variations in these properties, and thus the identification and study of "benchmark BDs" whose age and composition can be established independently. We identify the best sources of benchmark BDs as young open cluster members, moving group members, and wide (>1000AU) BD companions to both subgiant stars and high mass white dwarfs (WDs). We have used 2MASS to measure a wide L dwarf companion fraction of 2.7(+0.7/-0.5)%, which equates to a BD companion fraction of 34(+9/-6)% for an alpha~1 companion mass function. Using this value we simulate populations of wide BD binaries, and estimate that 80(+21/-14) subgiant--BD binaries, and 50(+13/-10) benchmark WD--BD binaries could be identified using current and new facilities. The WD--BD binaries should all be identifiable using the Large Area Survey component of UKIDSS combined with Sloan. Discovery of the subgiant--BD binaries will require a NIR imaging campaign around a large (~900) sample of Hipparcos subgiants. If identified, spectral studies of these benchmark brown dwarfs could reveal the spectral sensitivities across the Teff, g and [M/H] space probed by new surveys.