Understanding Stellar Velocity Measurements in Nearby Galaxies

Spectroscopy of the resolved stellar population of nearby galaxies yields line-of-sight velocities and chemical abundance information, and this can be used to study the internal kinematics of the stellar disks and halos of galaxies and their formation and evolution. Our research group used Keck DEIMOS multi-slit spectra of the resolved stellar populations of the Local Group galaxies M31, M32, and M33: a total of 81 multi-slit masks targeting ~18,000 stars. Despite the team’s best efforts to align the multi-slit mask prior to the spectroscopic observations, imperfect alignment of stars in multi-slit masks causes systematic error in velocity measurement. A relatively small amount of misalignment of the star within the slit (as small as 0.1" offset along the width direction of the slit) can cause a substantial systematic error in the velocity measurement. Fortunately, the so-called A-band at 7600A, one of the strongest telluric absorption features, allows for an after-the-fact measurement of misalignment of a star within the slit and correction of the corresponding velocity error (A-band correction). We studied mask-by-mask trends in the A-band correction versus position along the long axis of the DEIMOS mask and fit a third-order polynomial to the corrections of well measured stars. Linear trends and offsets indicate rotational and translational errors in the mask alignment, respectively, while curvature suggests imprecision in astrometry. Masks that used the 600 lines/mm grating show more scatter in the A-band correction than masks that used the 1200 lines/mm grating due to the worse spectral resolution of the former. These observed trends in the A-band correction are being used to improve the precision of stellar velocity measurements.