Identifying bright stars in crowded environments using velocity dispersion measurements, and an application to the center of M32

The identification of individual stars in crowded environments using photometric information alone is confounded by source confusion. However, with the addition of spectroscopic information it is possible to distinguish between blends and areas where the light is dominated by a single star, using the widths of absorption features. We describe a procedure for identifying locations in kinematically hot environments where the light is dominated by a single star, and apply this method to spectra with 0.1″ angular resolution covering the 2.1-2.3 μm interval in the central regions of M32. Targets for detailed investigation are selected as areas of localized brightness enhancement. Three locations where at least 60% of the K-band light comes from a single bright star, and another with light that is dominated by two stars with very different velocities, are identified. The dominant stars are evolving near the tip of the asymptotic giant branch (AGB), and have M5 III spectral type. The lack of a dispersion in spectral type suggests that the upper AGB within the central arcsecond of M32 has a dispersion in J - K of only a few hundredths of a magnitude, in agreement with what is seen at larger radii. One star has weaker atomic absorption lines than the others, such that [M/H] is 0.2 dex lower. Such a difference in metallicity is consistent with the metallicity dispersion inferred from the photometric width of the AGB in M32. The use of line width to distinguish between blends involving many relatively faint stars, none of which dominate the light output, and areas that are dominated by a single intrinsically bright star could be extended to crowded environments in other nearby galaxies.