Systematic uncertainties in stellar mass estimation for distinct galaxy populations

We show that different stellar mass estimation methods yield overall mass scales that disagree by factors up to similar to 2 for the z=0 galaxy population and, more importantly, relative mass scales that sometimes disagree by factors greater than or similar to 3 between distinct classes of galaxies (spiral/ irregular types, classical E/S0s, and E/S0s whose colors reflect recent star formation). This comparison considers stellar mass estimates based on (1) two different calibrations of the correlation between K-band mass-to-light ratio and B-R color and (2) detailed fitting of UBRJHK photometry and optical spectrophotometry using two different population synthesis models (Bruzual-Charlot and Maraston), with the same initial mass function in all cases. We also compare stellar+gas masses with dynamical masses. This analysis offers only weak arguments for preferring a particular stellar mass estimation method, given the plausibility of real variations in dynamical properties and dark matter content. These results help to calibrate the systematic uncertainties inherent in mass-based evolutionary studies of galaxies, including comparisons of low-and high-redshift galaxies.