The NIRSPEC Ultracool Dwarf Radial Velocity Survey

TL;DR

This infrared Doppler survey of ultracool dwarfs achieved high radial velocity precision, enabling detection of planetary companions and binary systems, and provided insights into their kinematic ages and binary fractions.

Contribution

The study demonstrates the effectiveness of telluric CH4 features for precise radial velocity measurements of ultracool dwarfs and reports the first orbital solutions for specific binary systems.

Findings

Achieved 50 m/s radial velocity precision for M dwarfs

Achieved ~200 m/s precision for L dwarfs

Estimated tight binary fraction of 2.5% in the sample

Abstract

We report the results of an infrared Doppler survey designed to detect brown dwarf and giant planetary companions to a magnitude-limited sample of ultracool dwarfs. Using the NIRSPEC spectrograph on the Keck II telescope, we obtained approximately 600 radial velocity measurements over a period of six years for a sample of 59 late-M and L dwarfs spanning spectral types M8/L0 to L6. A subsample of 46 of our targets have been observed on three or more epochs. We rely on telluric CH4 absorption features in the Earth's atmosphere as a simultaneous wavelength reference and exploit the rich set of CO absorption features found in the K-band spectra of cool stars and brown dwarfs to measure radial velocities and projected rotational velocities. For a bright, slowly rotating M dwarf standard we demonstrate a radial velocity precision of 50 m/s, and for slowly rotating L dwarfs we achieve a typical radial velocity precision of approximately 200 m/s. This precision is sufficient for the detection of close-in giant planetary companions to mid-L dwarfs as well as more equal mass spectroscopic binary systems with small separations (a<2 AU). We present an orbital solution for the subdwarf binary LSR1610-0040 as well as an improved solution for the M/T binary 2M0320-04. We also combine our radial velocity measurements with distance estimates and proper motions from the literature to estimate the dispersion of the space velocities of the objects in our sample. Using a kinematic age estimate we conclude that our UCDs have an age of 5.0+0.7-0.6 Gyr, similar to that of nearby sun-like stars. We simulate the efficiency with which we detect spectroscopic binaries and find that the rate of tight (a<1 AU) binaries in our sample is 2.5+8.6-1.6%, consistent with recent estimates in the literature of a tight binary fraction of 3-4%. (abridged)